rhetenor/awesome-awesomeness
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
Files
652812eed0d0ae17376945242133acb706f9b01f
awesome-awesomeness/html/agicocosci.md2.html
rhetenor 652812eed0 update
2025-07-18 23:13:11 +02:00

9583 lines
553 KiB
HTML
Raw Blame History

This file contains invisible Unicode characters
This file contains invisible Unicode characters that are indistinguishable to humans but may be processed differently by a computer. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
<div data-align="center">
<pre><code>&lt;img width=&quot;400&quot; height=&quot;253&quot; src=&quot;assets/abd_map.png&quot; alt=&quot;Roadmap of studying Abduction&quot;&gt;</code></pre>
</div>
<h1
id="awesome-artificial-general-intelligence-and-computational-cognitive-sciences-awesome">Awesome
Artificial General Intelligence and Computational Cognitive Sciences <a
href="https://awesome.re"><img src="https://awesome.re/badge.svg"
alt="Awesome" /></a></h1>
<p>An <strong>awesome &amp; curated</strong> list for <strong>Artificial
General Intelligence</strong>, an emerging inter-discipline field that
combines artificial intelligence and computational cognitive sciences as
majority, alone with probability and statistics, formal logic, cognitive
and developmental psychology, computational philosophy, cognitive
neuroscience, and computational sociology. We are promoting high-level
machine intelligence by getting inspirations from the way that human
learns and thinks, while obtaining a deeper understanding of human
cognition simultaneously. We believe that this kind of reciprocative
research is a potential way towards our big picture: building
human-level intelligent systems with capabilities such as abstracting,
explaining, learning, planning, and making decisions. And such
intelligence may generally help people improve scientific research,
engineering, and the arts, which are the hallmarks of human
intelligence.</p>
<p><strong><em>Awesome AGI &amp; CoCoSci</em></strong> is an all-in-one
collection, consisting of recources from basic courses and tutorials, to
papers and books around diverse topics in mutiple perspectives. Both
junior and senior researchers, whether learning, working on, or working
around AGI and CoCoSci, meet their interest here.</p>
<h2 id="contributing">Contributing</h2>
<p>Contributions are greatly welcomed! Please refer to <a
href="Contributing.md">Contribution Guidelines</a> before taking any
actions.</p>
<p><span id="c"></span> ## Contents</p>
<ul>
<li><a href="#papers">Papers</a>
<ul>
<li><a href="#abduction">Abduction</a>
<ul>
<li><a href="#explanation">Explanation</a></li>
<li><a href="#scientific-discovery">Scientific Discovery</a></li>
<li><a href="#rationalization">Rationalization</a></li>
<li><a href="#applications-in-ai">Applications in AI</a></li>
</ul></li>
<li><a href="#bayesian-modeling">Bayesian Modeling</a>
<ul>
<li><a href="#bayesian-induction">Bayesian Induction</a></li>
<li><a href="#generative-model">Generative Model</a></li>
<li><a href="#nonparametric-model">Nonparametric Model</a></li>
<li><a href="#bayesian-optimization">Bayesian Optimization</a></li>
</ul></li>
<li><a href="#concepts">Concepts</a>
<ul>
<li><a href="#theory-of-concepts">Theory of Concepts</a></li>
<li><a href="#human-concept-representation">Human Concept
Represenataion</a></li>
<li><a href="#ai-concept-representation">AI Concept
Representation</a></li>
</ul></li>
<li><a href="#complexity--information-theory">Complexity &amp;
Information Theory</a>
<ul>
<li><a href="#theory">Theory</a></li>
<li><a href="#dimensionality-reduction">Dimensionality
Reduction</a></li>
<li><a href="#visual-complexity">Visual Complexity</a></li>
</ul></li>
<li><a href="#communications">Communications</a>
<ul>
<li><a href="#non-verbal-communication">Non-Verbal
Communication</a></li>
<li><a href="#pragmatics">Pragmatics</a></li>
<li><a href="#language-compositionality">Language
Compositionality</a></li>
<li><a href="#coordination">Coordination</a></li>
</ul></li>
<li><a href="#domain-specific-language">Domain Specific Language</a>
<ul>
<li><a href="#design-theory">Design Theory</a></li>
<li><a href="#design-practises">Design Practises</a></li>
<li><a href="#design-automation">Design Automation</a></li>
<li><a href="#imperative-dsl-applications">Imperative DSL
Applications</a></li>
<li><a href="#declarative-dsl-applications">Declarative DSL
Applications</a></li>
<li><a href="#logic-dsl-applications">Logic DSL Applications</a></li>
<li><a href="#dsl-program-synthesis">DSL Program Synthesis</a></li>
<li><a href="#cognitive-foundations">Cognitive Foundations</a></li>
</ul></li>
<li><a href="#problem-solving">Problem Solving</a>
<ul>
<li><a href="#human-level-problem-solving">Human-Level Problem
Solving</a></li>
<li><a href="#planning">Planning</a></li>
<li><a href="#intrinsic-motivation">Intrinsic Motivation</a></li>
<li><a href="#reinforcement-learning">Reinforcement Learning</a></li>
<li><a href="#inverse-reinforcement-learning">Inverse Reinforcement
Learning</a></li>
</ul></li>
<li><a href="#system-1--system-2">System 1 &amp; System 2</a>
<ul>
<li><a href="#dual-coding-theory">Dual-Coding Theory</a></li>
<li><a href="#neural-symbolic-ai">Neural-Symbolic AI</a></li>
</ul></li>
<li><a href="#explainability">Explainability</a>
<ul>
<li><a href="#trustworthy-ai">Trustworthy AI</a></li>
<li><a href="#strong-machine-learning">Strong Machine Learning</a></li>
<li><a href="#explainable-deep-learning">Explainable Deep
Learning</a></li>
</ul></li>
<li><a href="#embodied-intelligence">Embodied Intelligence</a></li>
<li><a href="#evolutionary-intelligence">Evolutionary
Intelligence</a></li>
<li><a href="#methodologies-for-experiments">Methodologies for
Experiments</a>
<ul>
<li><a href="#quantitative-analysis">Quantitative Analysis</a></li>
<li><a href="#scaling-up-behavioral-studies">Scaling Up Behavioral
Studies</a></li>
<li><a href="#decision-making">Decision Making</a></li>
<li><a href="#question-answering">Question Answering</a></li>
<li><a href="#human-machine-comparison">Human-Machine
Comparison</a></li>
<li><a href="#association-test">Association Test</a></li>
<li><a href="#virtual-reality">Virtual Reality</a></li>
</ul></li>
<li><a href="#meta-level-considerations">Meta-Level Considerations</a>
<ul>
<li><a href="#meta-learning">Meta Learning</a></li>
<li><a href="#marrs-levels-of-analysis">Marr’s Levels of
Analysis</a></li>
<li><a href="#gestalt">Gestalt</a></li>
<li><a href="#the-aha-moment">The Aha! Moment</a></li>
<li><a href="#rationality">Rationality</a></li>
<li><a href="#cognitive-architecture">Cognitive Architecture</a></li>
</ul></li>
<li><a href="#science-logology">Science Logology</a>
<ul>
<li><a href="#philosophy-of-science">Philosophy of Science</a></li>
<li><a href="#science-of-science">Science of Science</a></li>
<li><a href="#literature-mining">Literature Mining</a></li>
<li><a href="#scientific-writing">Scientific Writing</a></li>
<li><a href="#science-education">Science Education</a></li>
<li><a href="#democratization-of-science">Democratization of
Science</a></li>
<li><a href="#laboratory-automation">Laboratory Automation</a></li>
<li><a href="#ai-assisted-research">AI Assisted Research</a></li>
</ul></li>
<li><a href="#theory-of-mind">Theory of Mind</a></li>
<li><a href="#analogy">Analogy</a></li>
<li><a href="#causality">Causality</a></li>
<li><a href="#commonsense">Commonsense</a>
<ul>
<li><a href="#intuitive-physics">Intuitive Physics</a></li>
<li><a href="#ai-commonsense-reasoning">AI Commonsense
Reasoning</a></li>
<li><a href="#commonsense-knowledgebase">Commonsense
Knowledgebase</a></li>
</ul></li>
<li><a href="#inductive-logic--program-synthesis">Inductive Logic &amp;
Program Synthesis</a></li>
<li><a href="#knowledge-representation">Knowledge
Representation</a></li>
<li><a href="#cognitive-development">Cognitive Development</a></li>
<li><a href="#learning-in-the-open-world">Learning in the Open
World</a></li>
<li><a href="#learning-with-cognitive-plausibility">Learning with
Cognitive Plausibility</a> <!--* [Tasks & Environments](#te)--></li>
</ul></li>
<li><a href="#academic-tools">Academic Tools</a>
<ul>
<li><a href="#courses">Courses</a></li>
<li><a href="#programming">Programming</a></li>
<li><a href="#paper-writing">Paper Writing</a></li>
<li><a href="#paper-reading">Paper Reading</a></li>
<li><a href="#literature-management">Literature Management</a></li>
<li><a href="#knowledge-management">Knowledge Management</a></li>
</ul></li>
<li><a href="#institute--researcher">Institute &amp; Researcher</a>
<ul>
<li><a href="#mit">MIT</a></li>
<li><a href="#stanford">Stanford</a></li>
<li><a href="#princeton">Princeton</a></li>
<li><a href="#harvard">Harvard</a></li>
<li><a href="#ucla">UCLA</a></li>
<li><a href="#uc-berkeley">UC Berkeley</a></li>
<li><a href="#bnu">BNU</a></li>
<li><a href="#pku">PKU</a></li>
<li><a href="#ucsd">UCSD</a></li>
<li><a href="#nyu">NYU</a></li>
<li><a href="#jhu">JHU</a></li>
<li><a href="#sit">SIT</a></li>
</ul></li>
<li><a href="#people--book">People &amp; Book</a>
<ul>
<li><a href="#john-hopcroft">John Hopcroft</a></li>
<li><a href="#ulf-grenander">Ulf Grenander</a></li>
<li><a href="#david-marr">David Marr</a></li>
<li><a href="#michael-tomasello">Michael Tomasello</a></li>
<li><a href="#judea-pearl">Judea Pearl</a></li>
<li><a href="#susan-carey">Susan Carey</a></li>
<li><a href="#daniel-kahneman">Daniel Kahneman</a></li>
<li><a href="#karl-popper">Karl Popper</a></li>
</ul></li>
<li><a href="#about">About</a></li>
</ul>
<h2 id="papers">Papers</h2>
<h3 id="abduction">Abduction</h3>
<h4 id="explanation">Explanation</h4>
<ul>
<li><p><a
href="https://plato.stanford.edu/entries/abduction/index.html">Abduction</a>
- <strong><em>Plato Stanford</em></strong>. A computational philosophy
account on Abduction, one of the three thinking patterns besides
Induction and Deduction, being unique for its potential to introduce new
ideas into current knowledge.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/scientific-explanation/">Scientific
Explanation</a> - <strong><em>Plato Stanford</em></strong>. A
computational philosophy account on Scientific Explanation, a canonical
application of Abduction.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/scientific-reduction/">Scientific
Reduction</a> - <strong><em>Plato Stanford</em></strong>. A
computational philosophy account on Scientific Reduction, which comes
with no explicit boundary with Explanation.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/logic-nonmonotonic/">Non-monotonic
Logic</a> - <strong><em>Plato Stanford</em></strong>. A computational
philosophy account on Non-monotonic Logic, a family of formal frameworks
devised to capture and represent defeasible inference.</p></li>
<li><p><a href="https://4lib.org/book/702071/e8ffe8">Philosophical
Writings of Peirce</a> - <strong><em>Courier Corporation</em></strong>,
1955. [<a
href="https://scholar.google.com/scholar?cluster=3917019015464129592">All
Versions</a>]. Original writings by C. S. Peirce, the philosopher who
first introduces the concept of Abduction.</p></li>
<li><p><a
href="https://www.hps.cam.ac.uk/files/lipton-inference.pdf">Inference to
the Best Explanation</a> - <strong><em>Routledge</em></strong>, 1991.
[<a
href="https://scholar.google.com/scholar?cluster=5097986614430666854">All
Versions</a>]. Lipton’s original paper on Inference to the Best
Explanation as a specialized condition of Abduction.</p></li>
<li><p><a
href="https://link.springer.com/book/10.1007/978-94-017-1733-5">Abductive
Reasoning and Learning</a> - <strong><em>Springer</em></strong>, 2000.
[<a
href="https://scholar.google.com/scholar?cluster=12074269365138058159">All
Versions</a>]. This book contains leading survey papers on the various
aspects of Abduction, both logical and numerical approaches.</p></li>
<li><p><a
href="https://link.springer.com/book/10.1007%2F978-3-642-03631-6">Abductive
Cognition: The Epistemological and Eco-Cognitive Dimensions of
Hypothetical Reasoning</a> - <strong><em>Springer</em></strong>, 2009.
[<a
href="https://scholar.google.com/scholar?cluster=8707351442527595188">All
Versions</a>]. Most philosophers of science in the twentieth century
have concluded that no logic of creative processes exists and, moreover,
that a rational model of discovery is impossible. In short, scientific
creative inferences are irrational and there is no “reasoning” to
hypotheses. On the other hand, some research in the area of artificial
intelligence has shown that methods for discovery could be found that
are computationally adequate for rediscovering — or discovering for the
first time — empirical or theoretical laws and theorems.</p></li>
<li><p><a
href="https://cognition.princeton.edu/sites/default/files/cognition/files/explanation_abductive_inference.pdf">Explanation
and Abductive Inference</a> - <strong><em>The Oxford Handbook of
Thinking and Reasoning</em></strong>, 2012. [<a
href="https://scholar.google.com/scholar?cluster=16126850654692681562">All
Versions</a>]. This chapter reviews evidence from cognitive psychology
and cognitive development concerning the structure and function of
explanations, with a focus on the role of explanations in learning and
inference. The findings highlight the value of understanding explanation
and abductive inference both as phenomena in their own right and for the
insights they provide concerning foundational aspects of human
cognition, such as representation, learning, and inference.</p></li>
<li><p><a
href="https://www.cell.com/AJHG/fulltext/S1364-6613(06)00132-X">Probabilistic
models of cognition: Conceptual foundations</a> - <strong><em>Trends in
Cognitive Sciences</em></strong>, 2006. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=12857321660837478492">All
Versions</a>]. Remarkable progress in the mathematics and computer
science of probability has led to a revolution in the scope of
probabilistic models. In particular, ‘sophisticated’ probabilistic
methods apply to structured relational systems such as graphs and
grammars, of immediate relevance to the cognitive sciences. This review
outlines progress in this rapidly developing field, which provides a
potentially unifying perspective across a wide range of domains and
levels of explanation.</p></li>
<li><p><a
href="https://cognition.princeton.edu/sites/default/files/cognition/files/tics_explanation.pdf">The
structure and function of explanations</a> - <strong><em>Trends in
Cognitive Sciences</em></strong>, 2006. [<a
href="https://scholar.google.com/scholar?cluster=2849189270394400667">All
Versions</a>]. Generating and evaluating explanations is spontaneous,
ubiquitous and fundamental to our sense of understanding. Recent
evidence suggests that in the course of an individual’s reasoning,
engaging in explanation can have profound effects on the probability
assigned to causal claims, on how properties are generalized and on
learning. These effects follow from two properties of the structure of
explanations: explanations accommodate novel information in the context
of prior beliefs, and do so in a way that fosters
generalization.</p></li>
<li><p><a
href="https://scholar.princeton.edu/sites/default/files/cognition/files/explanatory_prefs_tics.pdf">Explanatory
Preferences Shape Learning and Inference</a> - <strong><em>Trends in
Cognitive Sciences</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=2040551538203889465">All
Versions</a>]. People often learn by seeking explanations, and they
assess the viability of hypotheses by considering how well they explain
the data. An emerging body of work reveals that both children and adults
have strong and systematic intuitions about what constitutes a good
explanation, and that these explanatory preferences have a systematic
impact on explanation-based processes. In particular, people favor
explanations that are simple and broad, with the consequence that
engaging in explanation can shape learning and inference by leading
people to seek patterns and favor hypotheses that support broad and
simple explanations.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0010027715000955">The
Role of Explanatory Considerations in Updating</a> -
<strong><em>Cognition</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=3089358487428261042">All
Versions</a>]. This paper investigates experimentally controversy in
philosophy about the connection between explanation and inference, of
whether judgments of the explanatory goodness of hypotheses do play a
role when people revise their degrees of belief in those hypotheses upon
the receipt of new evidence.</p></li>
<li><p><a
href="https://www.tandfonline.com/doi/full/10.1080/20445911.2016.1230122">Explanation,
updating, and accuracy</a> - <strong><em>Journal of Cognitive
Psychology</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=967127146748155733">All
Versions</a>]. There is evidence that people update their credences
partly on the basis of explanatory considerations. Philosophers have
recently argued that to minimise the inaccuracy of their credences,
people’s updates also ought to be partly based on such considerations.
However, there are many ways in which explanatory considerations can
factor into updating, not all of which minimise inaccuracy. It is an
open question whether in their updating, people take explanatory
considerations into account in a way that philosophers would deem
recommendable. To address this question, the authors re-analyse data
from an experiment reported in Douven and Schupbach, “The role of
explanatory considerations in updating”.</p></li>
<li><p><a href="https://psycnet.apa.org/record/2018-03972-001">Best,
second-best, and good-enough explanations: How they matter to
reasoning</a> - <strong><em>Journal of Experimental
Psychology</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=3067550385175104201">All
Versions</a>]. There is a wealth of evidence that people’s reasoning is
influenced by explanatory considerations. Three experiments investigate
the descriptive adequacy of a precise proposal to be found in the
philosophical literature, to wit, that we should infer to the best
explanation, provided certain additional conditions are met. The main
conslusions are that (a) the quality of an explanation is a good
predictor of people’s willingness to accept that explanation, and a
better predictor than the prior probability of the explanation, and (b)
if more than one possible explanation is given, people are the less
willing to infer the best explanation the better they deem the
second-best explanation.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S1364661321001790">How
explanation guides belief change</a> - <strong><em>Trends in Cognitive
Sciences</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=15240531165875981526">All
Versions</a>]. Philosophers have argued that people ought to change
their graded beliefs via Bayes’ rule. Recent work in psychology
indicates that people sometimes violate that rule by attending to
explanatory factors. Results from computational modeling suggest that
such violations may actually be rational.</p></li>
<li><p><a
href="https://onlinelibrary.wiley.com/doi/abs/10.1207/s15516709cog2506_2">Use
of current explanations in multicausal abductive reasoning</a> -
<strong><em>Cognitive Science</em></strong>, 2001. [<a
href="https://scholar.google.com/scholar?cluster=7816050625957759346&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.pnas.org/content/110/42/16766.short">Kinematic mental
simulations in abduction and deduction</a> - <strong><em>Proceedings of
the National Academy of Sciences</em></strong>, 2013. [<a
href="https://scholar.google.com/scholar?cluster=11864820390497230588">All
Versions</a>]. This paper presents a theory, and its computer
implementation, of how mental simulations underlie the abductions of
informal algorithms and deductions from these algorithms. Three
experiments tested the theory’s predictions, using an environment of a
single railway track and a siding. The results corroborated the use of a
kinematic mental model in creating and testing informal algorithms and
showed that individuals differ reliably in the ability to carry out
these tasks.</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007/s11229-007-9223-4">Patterns
of abduction</a> - <strong><em>Synthese</em></strong>, 2007. [<a
href="https://scholar.google.com/scholar?cluster=15230540023076470385&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A categorization for Abduction in the account of pure
philosophy.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S1570868314000895">Abduction:
A categorical characterization</a> - <strong><em>Journal of Applied
Logic</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=17834260152484836885&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.journals.uchicago.edu/doi/abs/10.1086/392744">Defending
Abduction</a> - <strong><em>Philosophy of Science</em></strong>, 1999.
[<a
href="https://scholar.google.com/scholar?cluster=13895790050138832555&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007/s11229-009-9709-3">On
the distinction between Peirce’s abduction and Lipton’s Inference to the
best explanation</a> - <strong><em>Synthese</em></strong>, 2011. [<a
href="https://scholar.google.com/scholar?cluster=7865291004729010145&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007/s11229-019-02337-z">Abduction − the
context of discovery + underdetermination = inference to the best
explanation</a> - <strong><em>Synthese</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=4261649938116694095&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://link.springer.com/chapter/10.1007%2F3-540-45004-1_14">Towards
an Architecture for Cognitive Vision Using Qualitative Spatio-temporal
Representations and Abduction</a> - <strong><em>Spatial
Cognition</em></strong>, 2002. [<a
href="https://scholar.google.com/scholar?cluster=8072265283930278310&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007/s11229-018-1824-6">Abductive
inference within a pragmatic framework</a> -
<strong><em>Synthese</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=10285954503043361393&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007/s00354-019-00059-x">Disjunctive
Abduction</a> - <strong><em>New Generation Computing</em></strong>,
2019. [<a
href="https://scholar.google.com/scholar?cluster=6664745483675209831&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.frontiersin.org/articles/10.3389/fpsyg.2015.00459/full">Probabilistic
alternatives to Bayesianism: the case of explanationism</a> -
<strong><em>Frontiers in Psychology</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=9016714668469830914&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A non-Bayesian account of Abduction.</p></li>
<li><p><a
href="https://www.scitepress.org/Link.aspx?doi=10.5220/0010195405620571">A
Probabilistic Theory of Abductive Reasoning</a> -
<strong><em>ICAART</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=450937566244876051&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A probabilistic perspective for interpreting Abductive
Reasoning.</p></li>
<li><p><a
href="https://www.tandfonline.com/doi/full/10.1080/09528130600558141?scroll=top&amp;needAccess=true">The
order effect in human abductive reasoning: an empirical and
computational study</a> - <strong><em>Journal of Experimental &amp;
Theoretical Artificial Intelligence</em></strong>, 2006. [<a
href="https://scholar.google.com/scholar?cluster=3803536062463585043&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://link.springer.com/chapter/10.1007%2F978-3-642-15223-8_5">Abduction,
Induction, and Analogy</a> - <strong><em>Model-Based Reasoning in
Science and Technology</em></strong>, 2010. [<a
href="https://scholar.google.com/scholar?cluster=14979764682921693390&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The distinctions and relations between Abduction,
Induction, and Analogy.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/abs/pii/S0010027718301094">Remembrance
of inferences past: Amortization in human hypothesis generation</a> -
<strong><em>Cognition</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=190340622765037472&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>]. A rational account of human hypothesis
generation.</p></li>
<li><p><a
href="https://onlinelibrary.wiley.com/doi/full/10.1111/j.1551-6709.2010.01142.x">The
AHA! Experience: Creativity Through Emergent Binding in Neural
Networks</a> - <strong><em>Cognitive Science</em></strong>, 2012. [<a
href="https://scholar.google.com/scholar?cluster=10006889101167052798&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S2352154620300851">Explanation-seeking
curiosity in childhood</a> - <strong><em>Current Opinion in Behavioral
Sciences</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=4167956555501133663&amp;hl=en&amp;as_sdt=2005">All
Versions</a>]. A piece of developmental pshchological evidence for
Abduction in young children.</p></li>
<li><p><a href="https://arxiv.org/abs/2204.14267">A Grammar of
Hypotheses for Visualization, Data, and Analysis</a> - 2022. [<a
href="https://scholar.google.com/scholar?cluster=10321469321980973246">All
Versions</a>]. This work presents a grammar for expressing hypotheses in
visual data analysis to formalize the previously abstract notion of
“analysis tasks.” Through the lens of this grammar, the authors lay the
groundwork for how a user’s data analysis questions can be
operationalized and automated as a set of hypotheses (a hypothesis
space). The authors demonstrate that the grammar-based approach for
analysis tasks can provide a systematic method towards unifying three
disparate spaces in visualization research: the hypotheses a dataset can
express (a data hypothesis space), the hypotheses a user would like to
refine or verify through analysis (an analysis hypothesis space), and
the hypotheses a visualization design is capable of supporting (a
visualization hypothesis space). The authors illustrate how the
formalization of these three spaces can inform future research in
visualization evaluation, knowledge elicitation, analytic provenance,
and visualization recommendation by using a shared language for
hypotheses. Finally, the authors compare the proposed grammar-based
approach with existing visual analysis models and discuss the potential
of a new hypothesis-driven theory of visual analytics.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="scientific-discovery">Scientific Discovery</h4>
<ul>
<li><p><a
href="https://plato.stanford.edu/entries/scientific-discovery/">Scientific
Discovery</a> - <strong><em>Plato Stanford</em></strong>. A
computational philosophy account on Scientific Discovery, the process or
product of successful scientific inquiry, sometimes an Abduction-like
(Explanation) thinking pattern.</p></li>
<li><p><a
href="https://hk1lib.org/book/2241843/c5d7b3?id=2241843&amp;secret=c5d7b3">Models
of Discovery: And Other Topics in the Methods of Science</a> -
<strong><em>Springer</em></strong>, 1977. [<a
href="https://scholar.google.com/scholar?cluster=9932701864897299105&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original book on search as scientific
thinking.</p></li>
<li><p><a
href="https://hk1lib.org/book/970300/6b0ff7?id=970300&amp;secret=6b0ff7">Scientific
discovery: Computational explorations of the creative processes</a> -
<strong><em>MIT Press</em></strong>, 1987. [<a
href="https://scholar.google.com/scholar?cluster=11327000316248254911">All
Versions</a>]. The book is divided into four parts. Part I introduces
the subject of discovery, defines the scope of our work, and discusses
some of the issues that have surrounded and still surround our topic.
Parts II and III contain the main body of our results, largely in the
form of accounts of the performance of computer programs that simulate
human thought processes to make scientific discoveries. Part II is
devoted largely to the processes for inducing quantitative theories from
data. Part III is devoted mainly to the processes for inducing
qualitative descriptive and structural theories from data. In Part IV,
on the basis of our experience, we discuss at a lower level of precision
how the programs described in the preceding chapters could be combined
into a single, more general discovery system, and we describe a wide
range of the other component processes that enter into scientific
discovery.</p></li>
<li><p><a href="https://psycnet.apa.org/record/2000-03968-000">Exploring
science: The cognition and development of discovery processes</a> -
<strong><em>MIT Press</em></strong>, 2000. [<a
href="https://scholar.google.com/scholar?cluster=13091264356550286420">All
Versions</a>]. In this book, D. Klahr sets out to describe the cognitive
and developmental processes that have enabled scientists to make the
discoveries that comprise the body of information we call “scientific
knowledge.” Over the past decade, Klahr and his colleagues have
conducted laboratory experiments in which they create discovery
contexts, computer-based environments, to evoke the kind of thinking
characteristic of scientific discovery in the “real world.” In
attempting to solve the problems posed by the discovery tasks,
experiment participants (from preschoolers to university students) use
many of the same higher-order cognitive processes used by practicing
scientists. Through his work, Klahr integrates two disparate
approaches–the content-based approach and the process-based approach– to
present a comprehensive model of the psychology of scientific
discovery.</p></li>
<li><p><a
href="https://onlinelibrary.wiley.com/doi/abs/10.1207/s15516709cog1201_1">Dual
Space Search During Scientific Reasoning</a> - <strong><em>Cognitive
Science</em></strong>, 1988. [<a
href="https://scholar.google.com/scholar?cluster=17542852673494089523&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>]. The original paper on the dual space search as scientific
thinking theory.</p></li>
<li><p><a href="https://escholarship.org/uc/item/94n547fj">Complexity
Management in a Discovery Task</a> -
<strong><em>CogSci’92</em></strong>, 1992. [<a
href="https://scholar.google.com/scholar?cluster=18138712608977258974">All
Versions</a>]. Previous psychological research about scientific
discovery has often focused on subjects’ heuristics for discovering
simple concepts with one relevant dimension or a few relevant dimensions
with simple two-way interactions. This paper presents results from an
experiment in which subjects had to discover a concept involving complex
three-way interactions on a multi-valued output by running experiments
in a computerized microworld. Twenty-two CMU undergraduates attempted
the task, of which sixteen succeeded, in an average of 85 minutes. The
analyses focus on three strategies used to regulate task complexity.
First, subjects preferred depth-first to breadth-first search, with
successful subjects regulating the number of features varied from
experiment to experiment most effectively. Second, subjects
systematically regulated the length of their experiments. Third, a new
explicit search heuristic (Put Upon Stack Heuristic) used by successful
subjects is described.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S1071581996900324">A
dual-space model of iteratively deepening exploratory learning</a> -
<strong><em>International Journal of Human-Computer
Studies</em></strong>, 1996. [<a
href="https://scholar.google.com/scholar?cluster=17337189265334825678">All
Versions</a>]. This paper describes a cognitive model of exploratory
learning, which covers both trial-and-error and instruction-taking
activities. The model, implemented in Soar, is grounded in empirical
data of subjects in a task-oriented, trial-and-error exploratory
learning situation. A key empirical finding reflected in the model is
the repeated scanning of a subset of the available menu items, with
increased attention to items on each successive scan. This is explained
in terms of dual search spaces, the external interface and the user’s
internal knowledge, both of which must be tentatively explored with
attention to changing costs and benefits.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/abs/pii/S0010028583710030">Heuristics
for Scientific Experimentation: A Developmental Study</a> -
<strong><em>Cognitive Psychology</em></strong>, 1993. [<a
href="https://scholar.google.com/scholar?cluster=2469515962071844494&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>]. A piece of evidence on children have basic scientific
thinking skills.</p></li>
<li><p><a
href="https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.645.248&amp;rep=rep1&amp;type=pdf">A
4-Space Model of Scientific Discovery</a> -
<strong><em>CogSci’95</em></strong>, 1995. [<a
href="https://scholar.google.com/scholar?cluster=1063157789682040473&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>]. Extending the dual space search.</p></li>
<li><p><a
href="https://link.springer.com/article/10.3758/BF03201090">When to
trust the data: Further investigations of system error in a scientific
reasoning task</a> - <strong><em>Memory &amp; Cognition</em></strong>,
1996. [<a
href="https://scholar.google.com/scholar?cluster=3131191372086488656">All
Versions</a>]. When evaluating experimental evidence, how do people deal
with the possibility that some of the feedback is erroneous? The
potential for error means that evidence evaluation must include
decisions about when to “trust the data.” This paper presents two
studies that focus on subjects’ responses to erroneous feedback in a
hypothesis testing situation—a variant of Wason’s (1960) 2–4–6 rule
discovery task in which some feedback was subject tosystem error: “hits”
were reported as “misses” and vice versa. Results show that, in contrast
to previous research, people are equally adept at identifying false
negatives and false positives; further, successful subjects were less
likely to use a positive test strategy (Klayman &amp; Ha, 1987) than
were unsuccessful subjects. Finally, although others have found that
generating possible hypotheses prior to experimentation increases
success and task efficiency, such a manipulation did little to mitigate
the effects of system error.</p></li>
<li><p><a
href="https://psycnet.apa.org/record/1987-20689-001">Confirmation,
disconfirmation, and information in hypothesis testing</a> -
<strong><em>Psychological Review</em></strong>, 1987. [<a
href="https://scholar.google.com/scholar?cluster=1954141597807453515&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A psychological account on hypothesis testing.</p></li>
<li><p><a
href="https://psycnet.apa.org/record/2010-22980-001">Hypothesis
generation, sparse categories, and the positive test strategy</a> -
<strong><em>Psychological Review</em></strong>, 2011. [<a
href="https://scholar.google.com/scholar?cluster=4329636480235863472&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://psycnet.apa.org/record/1990-03504-001">Children
and adults as intuitive scientists</a> - <strong><em>Psychological
Review</em></strong>, 1989. [<a
href="https://scholar.google.com/scholar?cluster=9577945454476127070&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>]. A perspective against search as scientific
thinking.</p></li>
<li><p><a
href="https://link.springer.com/content/pdf/10.1007/s11229-019-02127-7.pdf">Abduction
and styles of scientific thinking</a> -
<strong><em>Synthese</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=9336871656706514469&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A computational philosophy account connecting Abduction
and scientific thinking.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="rationalization">Rationalization</h4>
<ul>
<li><p><a
href="https://www.sciencedirect.com/science/article/abs/pii/S0885201414000744">Imagination
and the generation of new ideas</a> - <strong><em>Cognitive
Development</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=16920774374067505248">All
Versions</a>]. A variety of theories have been put forth to explain the
function of imagination, most notably that imagination engages and
develops children’s theory of mind and counterfactual reasoning. This
work proposes that a primary role for imagination is as a cognitive
mechanism for efficiently generating new ideas without observing new
evidence. Learners must generate hypotheses before they can assess the
truth of these hypotheses. Given infinite possibilities, how do learners
constrain the process of hypothesis generation? The authors suggest that
learners represent abstract criteria for the solution to a problem and
generate solutions that, if true, would solve the problem. As a
preliminary test of this idea, the authors show that, in the absence of
any fact of the matter (i.e., when neither prior knowledge nor
statistical data distinguishes competing hypotheses), 4–6-year-olds
(mean: 63 months) systematically converge on solutions to problems,
consistent with an ability to imagine the abstract properties of causal
problems and their solutions.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S1364661319302311">How
We Know What Not To Think</a> - <strong><em>Trends in Cognitive
Sciences</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=13106919756521743226">All
Versions</a>]. Humans often represent and reason about unrealized
possible actions—the vast infinity of things that were not (or have not
yet been) chosen. This capacity is central to the most impressive of
human abilities: causal reasoning, planning, linguistic communication,
moral judgment, etc. Nevertheless, how do we select possible actions
that are worth considering from the infinity of unrealized actions that
are better left ignored? This work reviews research across the cognitive
sciences, and find that the possible actions considered by default are
those that are both likely to occur and generally valuable. This paper
then offers a unified theory of why. The authors propose that (i) across
diverse cognitive tasks, the possible actions we consider are biased
towards those of general practical utility, and (ii) a plausible primary
function for this mechanism resides in decision making.</p></li>
<li><p><a
href="https://www.cambridge.org/core/journals/behavioral-and-brain-sciences/article/abs/rationalization-is-rational/2A13B99ED09BD802C0924D3681FEC55B">Rationalization
is rational</a> - <strong><em>Behavioral and Brain
Sciences</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=5165464589274056844">All
Versions</a>]. [<a
href="https://bpb-us-e1.wpmucdn.com/websites.harvard.edu/dist/0/59/files/2022/03/rationalization_is_rational.pdf">Preprint</a>].
Rationalization occurs when a person has performed an action and then
concocts the beliefs and desires that would have made it rational. Then,
people often adjust their own beliefs and desires to match the concocted
ones. While many studies demonstrate rationalization, and a few theories
describe its underlying cognitive mechanisms, we have little
understanding of its function. Why is the mind designed to construct
post hoc rationalizations of its behavior, and then to adopt them? This
may accomplish an important task: transferring information between the
different kinds of processes and representations that influence our
behavior. Human decision making does not rely on a single process; it is
influenced by reason, habit, instinct, norms, and so on. Several of
these influences are not organized according to rational choice (i.e.,
computing and maximizing expected value). Rationalization extracts
implicit information – true beliefs and useful desires – from the
influence of these non-rational systems on behavior.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S1364661321001480">Rationalizing
constraints on the capacity for cognitive control</a> -
<strong><em>Trends in Cognitive Sciences</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=13060297961922073063">All
Versions</a>]. Humans are remarkably limited in: (i) how many
control-dependent tasks they can execute simultaneously, and (ii) how
intensely they can focus on a single task. These limitations are
universal assumptions of most theories of cognition. Yet, a rationale
for why humans are subject to these constraints remains elusive. This
feature review draws on recent insights from psychology, neuroscience,
and machine learning, to suggest that constraints on cognitive control
may result from a rational adaptation to fundamental, computational
dilemmas in neural architectures. The reviewed literature implies that
limitations in multitasking may result from a trade-off between learning
efficacy and processing efficiency and that limitations in the intensity
of commitment to a single task may reflect a trade-off between cognitive
stability and flexibility.</p></li>
<li><p><a
href="https://www.cambridge.org/core/journals/behavioral-and-brain-sciences/article/abs/why-imaginary-worlds/CA2AB4B1E1EDD8FE965E6DDB4A047B35">Why
Imaginary Worlds? The psychological foundations and cultural evolution
of fictions with imaginary worlds</a> - <strong><em>Behavioral and Brain
Sciences</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=16985691366494688837">All
Versions</a>]. Imaginary worlds are extremely successful. The most
popular fictions produced in the last few decades contain such a
fictional world. They can be found in all fictional media, from novels
(e.g., Lord of The Rings and Harry Potter) to films (e.g., Star Wars and
Avatar), video games (e.g., The Legend of Zelda and Final Fantasy),
graphic novels (e.g., One Piece and Naruto), and TV series (e.g., Star
Trek and Game of Thrones), and they date as far back as ancient
literature (e.g., the Cyclops Islands in The Odyssey, 850 BCE). Why such
a success? Why so much attention devoted to non-existent worlds? In this
paper, the authors propose that imaginary worlds co-opt our preferences
for exploration, which have evolved in humans and nonhuman animals
alike, to propel individuals toward new environments and new sources of
reward. Humans would find imaginary worlds very attractive for the very
same reasons, and under the same circumstances, as they are lured by
unfamiliar environments in real life. After reviewing research on
exploratory preferences in behavioral ecology, environmental esthetics,
neuroscience, and evolutionary and developmental psychology, the authors
focus on the sources of their variability across time and space, which
they argue can account for the variability of the cultural preference
for imaginary worlds. This hypothesis can, therefore, explain the way
imaginary worlds evolved culturally, their shape and content, their
recent striking success, and their distribution across time and
populations.</p></li>
<li><p><a href="https://escholarship.org/uc/item/5f64z7d7">Coalescing
the Vapors of Human Experience into a Viable and Meaningful
Comprehension</a> - <strong><em>CogSci’16</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=5460385008324352958">All
Versions</a>]. Models of concept learning and theory acquisition often
invoke a stochastic search process, in which learners generate
hypotheses through some structured random process and thenevaluate them
on some data measuring their quality or value. To be successful within a
reasonable time-frame, these models need ways of generating good
candidate hypotheses evenbefore the data are considered. Schulz (2012a)
has proposed that studying the origins of new ideas in more everyday
contexts, such as how we think up new names for things, can provide
insight into the cognitive processes that generate good hypotheses for
learning. We propose a simple generative model for how people might draw
on their experience to propose new names in everyday domains such as pub
names or action movies, and show that it captures surprisingly well the
names that people actually imagine. We discuss the role for an analogous
hypothesis-generation mechanism in enabling and constraining causal
theory learning.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="applications-in-ai">Applications in AI</h4>
<ul>
<li><p><a href="https://www.nature.com/articles/nature02236">Functional
genomic hypothesis generation and experimentation by a robot
scientist</a> - <strong><em>Nature</em></strong>, 2004. [<a
href="https://scholar.google.com/scholar?cluster=17461972625475533182">All
Versions</a>]. This paper describes a physically implemented robotic
system that applies techniques from artificial intelligence to carry out
cycles of scientific experimentation. The system automatically
originates hypotheses to explain observations, devises experiments to
test these hypotheses, physically runs the experiments using a
laboratory robot, interprets the results to falsify hypotheses
inconsistent with the data, and then repeats the cycle. The system is
applied to the determination of gene function using deletion mutants of
yeast (Saccharomyces cerevisiae) and auxotrophic growth experiments. The
authors built and tested a detailed logical model (involving genes,
proteins and metabolites) of the aromatic amino acid synthesis
pathway.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/abs/pii/0004370293900154?via%3Dihub">Interpretation
as abduction</a> - <strong><em>Artificial Intelligence</em></strong>,
1993. [<a
href="https://scholar.google.com/scholar?cluster=12658433318211361322">All
Versions</a>]. Abduction is inference to the best explanation. The
authors have developed an approach to abductive inference, called
“weighted abduction”, that has resulted in a significant simplification
of how the problem of interpreting texts is conceptualized. The
interpretation of a text is the minimal explanation of why the text
would be true. More precisely, to interpret a text, one must prove the
logical form of the text from what is already mutually known, allowing
for coercions, merging redundancies where possible, and making
assumptions where necessary. It is shown how such “local pragmatics”
problems as reference resolution, the interpretation of compound
nominals, the resolution of syntactic ambiguity and metonymy, and schema
recognition can be solved in this manner. Moreover, this approach of
“interpretation as abduction” can be combined with the older view of
“parsing as deduction” to produce an elegant and thorough integration of
syntax, semantics, and pragmatics, one that spans the range of
linguistic phenomena from phonology to discourse structure.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/abs/pii/000437029390061F?via%3Dihub">Probabilistic
Horn abduction and Bayesian networks</a> - <strong><em>Artificial
Intelligence</em></strong>, 1993. [<a
href="https://scholar.google.com/scholar?cluster=7728248035489349629">All
Versions</a>]. This paper presents a simple framework for Horn-clause
abduction, with probabilities associated with hypotheses. The framework
incorporates assumptions about the rule base and independence
assumptions amongst hypotheses. It is shown how any probabilistic
knowledge representable in a discrete Bayesian belief network can be
represented in this framework. The main contribution is in finding a
relationship between logical and probabilistic notions of evidential
reasoning. This provides a useful representation language in its own
right, providing a compromise between heuristic and epistemic
adequacy.</p></li>
<li><p><a
href="https://link.springer.com/chapter/10.1007/978-3-540-39879-0_6">Abductive
Inference in Bayesian Networks: A Review</a> - <strong><em>Advances in
Bayesian Networks</em></strong>, 2004. [<a
href="https://scholar.google.com/scholar?cluster=8502276402734843212">All
Versions</a>]. The goal of this paper is to serve as a survey for the
problem of abductive inference (or belief revision) in Bayesian
networks. Thus, the problem is introduced in its two variants: total
abduction (or MPE) and partial abduction (or MAP) . Also, the problem is
formulated in its general case, that is, looking for the K best
explanations. Then, a (non exhaustive) review of exact and approximate
algorithms for dealing with both abductive inference problems is carried
out. Finally, the authors collect the main complexity results appeared
in the literature for both problems (MPE and MAP).</p></li>
<li><p><a
href="https://academic.oup.com/logcom/article-abstract/2/6/719/942121">Abductive
Logic Programming</a> - <strong><em>Journal of Logic
Computation</em></strong>, 1992. [<a
href="https://scholar.google.com/scholar?cluster=18119357517656745518">All
Versions</a>]. This paper is a survey and critical overview of recent
work on the extension of logic programming to perform abductive
reasoning (abductive logic programming). The authors outline the general
framework of abduction and its applications to knowledge assimilation
and default reasoning; and they introduce an argumentation-theoretic
approach to the use of abduction as an interpretation for negation as
failure.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0743106699000758">ACLP:
Abductive Constraint Logic Programming</a> - <strong><em>The Journal of
Logic Programming</em></strong>, 1999. [<a
href="https://scholar.google.com/scholar?cluster=14319574550421192429">All
Versions</a>]. This paper presents the framework of Abductive Constraint
Logic Programming (ACLP), which integrates Abductive Logic Programming
(ALP) and Constraint Logic Programming (CLP). In ACLP, the task of
abduction is supported and enhanced by its non-trivial integration with
constraint solving. This integration of constraint solving into
abductive reasoning facilitates a general form of constructive abduction
and enables the application of abduction to computationally demanding
problems. The paper studies the formal declarative and operational
semantics of the ACLP framework together with its application to various
problems.</p></li>
<li><p><a
href="https://link.springer.com/chapter/10.1007/3-540-45628-7_16">Abduction
in Logic Programming</a> - <strong><em>Computational
Logic</em></strong>, 2002. [<a
href="https://scholar.google.com/scholar?cluster=902643678163312237">All
Versions</a>]. [<a
href="https://web.stanford.edu/class/cs227/Readings/Abudction%20in%20LP.pdf">Preprint</a>].
Abduction in Logic Programming started in the late 80s, early 90s, in an
attempt to extend logic programming into a framework suitable for a
variety of problems in Artificial Intelligence and other areas of
Computer Science. This paper aims to chart out the main developments of
the field over the last ten years and to take a critical view of these
developments from several perspectives: logical, epistemological,
computational and suitability to application. The paper attempts to
expose some of the challenges and prospects for the further development
of the field.</p></li>
<li><p><a
href="https://dl.acm.org/doi/abs/10.5555/2283696.2283887">Bayesian
Abductive Logic Programs: A Probabilistic Logic for Abductive
Reasoning</a> - <strong><em>IJCAI’11</em></strong>, 2011. [<a
href="https://scholar.google.com/scholar?cluster=4453424083730209198">All
Versions</a>]. [<a
href="https://www.cs.utexas.edu/~ml/papers/raghavan.starai10.pdf">Preprint</a>].
This work introduces Bayesian Abductive Logic Programs (BALP), a
probabilistic logic that adapts Bayesian Logic Programs (BLPs) for
abductive reasoning. Like BLPs, BALPs also combine first-order logic and
Bayes nets. However, unlike BLPs, which use deduction to construct Bayes
nets, BALPs employ logical abduction. As a result, BALPs are more suited
for problems like plan/activity recognition that require abductive
reasoning.</p></li>
<li><p><a
href="https://www.cs.utexas.edu/~ml/papers/raghavan.ecml11.pdf">Abductive
Plan Recognition by Extending Bayesian Logic Programs</a> -
<strong><em>ECML’11</em></strong>, 2011. [<a
href="https://scholar.google.com/scholar?cluster=7276511797197017483">All
Versions</a>]. Plan recognition is the task of predicting an agent’s
top-level plans based on its observed actions. It is an abductive
reasoning task that involves inferring cause from effect. Most existing
approaches to plan recognition use either first-order logic or
probabilistic graphical models. While the former cannot handle
uncertainty, the latter cannot handle structured representations. In
order to overcome these limitations, this work develops an approach to
plan recognition using Bayesian Logic Programs (BLPs), which combine
first-order logic and Bayesian networks. Since BLPs employ logical
deduction to construct the networks, they cannot be used effectively for
plan recognition. Therefore, the authors extend BLPs to use logical
abduction to construct Bayesian networks and call the resulting model
Bayesian Abductive Logic Programs (BALPs). The authors learn the
parameters in BALPs using the Expectation Maximization algorithm adapted
for BLPs. Finally, the authors present an experimental evaluation of
BALPs on three benchmark data sets and compare its performance with the
state-of-the-art for plan recognition.</p></li>
<li><p><a
href="https://www.aaai.org/ocs/index.php/IJCAI/IJCAI13/paper/view/6624/6619">An
Approach to Abductive Reasoning in Equational Logic</a> -
<strong><em>IJCAI’13</em></strong>, 2013. [<a
href="https://scholar.google.com/scholar?cluster=686895264429811190">All
Versions</a>]. Abduction has been extensively studied in propositional
logic because of its many applications in artificial intelligence.
However, its intrinsic complexity has been a limitation to the
implementation of abductive reasoning tools in more expressive logics.
The authors have devised such a tool in ground flat equational logic, in
which literals are equations or disequations between constants. The tool
is based on the computation of prime implicates. It uses a relaxed
paramodulation calculus, designed to generate all prime implicates of a
formula, together with a carefully defined data structure storing the
implicates and able to efficiently detect, and remove,
redundancies.</p></li>
<li><p><a
href="https://ojs.aaai.org//index.php/AAAI/article/view/3964">Abduction-Based
Explanations for Machine Learning Models</a> -
<strong><em>AAAI’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=7355960657107994022">All
Versions</a>]. The growing range of applications of Machine Learning
(ML) in a multitude of settings motivates the ability of computing small
explanations for predictions made. Small explanations are generally
accepted as easier for human decision makers to understand. Most earlier
work on computing explanations is based on heuristic approaches,
providing no guarantees of quality, in terms of how close such solutions
are from cardinality- or subset-minimal explanations. This paper
develops a constraint-agnostic solution for computing explanations for
any ML model. The proposed solution exploits abductive reasoning, and
imposes the requirement that the ML model can be represented as sets of
constraints using some target constraint reasoning system for which the
decision problem can be answered with some oracle. The experimental
results, obtained on well-known datasets, validate the scalability of
the proposed approach as well as the quality of the computed
solutions.</p></li>
<li><p><a
href="https://www.ijcai.org/proceedings/2021/0424.pdf">Probabilistic
Sufficient Explanations</a> - <strong><em>IJCAI’21</em></strong>, 2021.
[<a
href="https://scholar.google.com/scholar?cluster=1874102360688341104">All
Versions</a>]. Understanding the behavior of learned classifiers is an
important task, and various black-box explanations, logical reasoning
approaches, and model-specific methods have been proposed. This paper
introduces probabilistic sufficient explanations, which formulate
explaining an instance of classification as choosing the “simplest”
subset of features such that only observing those features is
“sufficient” to explain the classification. That is, sufficient to give
us strong probabilistic guarantees that the model will behave similarly
when all features are observed under the data distribution. In addition,
the authors leverage tractable probabilistic reasoning tools such as
probabilistic circuits and expected predictions to design a scalable
algorithm for finding the desired explanations while keeping the
guarantees intact. The experiments demonstrate the effectiveness of the
algorithm in finding sufficient explanations, and showcase its
advantages compared to Anchors and logical explanations.</p></li>
<li><p><a href="https://www.aclweb.org/anthology/H91-1024.pdf">Machine
Translation Using Abductive Inference</a> -
<strong><em>COLING</em></strong>, 1990. [<a
href="https://scholar.google.com/scholar?cluster=15275163177548183539">All
Versions</a>]. Many existing approaches to machine translation take for
granted that the information presented in the output is found somewhere
in the input, and, moreover, that such information should be expressed
at a single representational level, say, in terms of the parse trees or
of “semantic” assertions. Languages, however, not only express the
equivalent information by drastically different linguistic means, but
also often disagree in what distinctions should be expressed
linguistically at all. For example, in translating from Japanese to
English, it is often necessary to supply determiners for noun phrases,
and this in general cannot be done without deep understanding of the
source text. Similarly, in translating from English to Japanese,
politeness considerations, which in English are implicit in the social
situation and explicit in very diffuse ways in, for example, the heavy
use of hypotheticals, must be realized grammatically in Japanese.
Machine translation therefore requires that the appropriate inferences
be drawn and that the text be interpreted to some depth. Recently, an
elegant approach to inference in discourse interpretation has been
developed at a number of sites, all based on the notion of abduction,
and the authors have begun to explore its potential application to
machine translation. The authors argue that this approach provides the
possibility of deep reasoning and of mapping between the languages at a
variety of levels.</p></li>
<li><p><a href="https://arxiv.org/abs/2105.07758">Automated Biodesign
Engineering by Abductive Meta-Interpretive Learning</a> -
<strong><em>AAAI Spring Symposium Series 2021 on Artificial Intelligence
for Synthetic Biology</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=543730388062329581">All
Versions</a>]. This work proposes an automated biodesign engineering
framework empowered by Abductive Meta-Interpretive Learning (MetaAbd), a
novel machine learning approach that combines symbolic and sub-symbolic
machine learning, to further enhance the design-build-test-learn cycle
by enabling the learning machine to 1) exploit domain knowledge and
learn human-interpretable models that are expressed by formal languages
such as first-order logic; 2) simultaneously optimise the structure and
parameters of the models to make accurate numerical predictions; 3)
reduce the cost of experiments and effort on data annotation by actively
generating hypotheses and examples.</p></li>
<li><p><a href="https://arxiv.org/abs/2308.12740">Human Comprehensible
Active Learning of Genome-Scale Metabolic Networks</a> -
<strong><em>AAAI Spring Symposium Series 2023 on Computational
Scientific Discovery</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=10875437066608527790">All
Versions</a>]. [<a
href="http://cogsys.org/symposium/discovery-2023/abstracts/Abstract_3169.pdf">Extended
Abstract</a>]. [<a
href="http://cogsys.org/symposium/discovery-2023/talks/Ai.pdf">Slides</a>].
This work introduces a novel machine learning framework ILP-iML1515
based on Inductive Logic Programming (ILP) that performs abductive
logical reasoning and actively learns from training examples. The
ILP-iML1515 framework 1) allows high-throughput simulations and 2)
actively selects experiments that reduce the experimental cost of
learning gene functions in comparison to randomly selected
experiments.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="bayesian-modeling">Bayesian Modeling</h3>
<h4 id="bayesian-induction">Bayesian Induction</h4>
<ul>
<li><p><a
href="https://plato.stanford.edu/entries/epistemology-bayesian/">Bayesian
Epistemology</a> - <strong><em>Plato Stanford</em></strong>. A
computational philosophy account on the nature of uncertainty modeling
in Bayesian Epistemology.</p></li>
<li><p><a
href="https://www.nature.com/articles/nature14541">Probabilistic machine
learning and artificial intelligence</a> -
<strong><em>Nature</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=1783282361269717744">All
Versions</a>]. Probabilistic modelling provides a framework for
understanding what learning is, and has therefore emerged as one of the
principal theoretical and practical approaches for designing machines
that learn from data acquired through experience. The probabilistic
framework, which describes how to represent and manipulate uncertainty
about models and predictions, has a central role in scientific data
analysis, machine learning, robotics, cognitive science and artificial
intelligence. This Review provides an introduction to this framework,
and discusses some of the state-of-the-art advances in the field,
namely, probabilistic programming, Bayesian optimization, data
compression and automatic model discovery.</p></li>
<li><p><a
href="https://www.cambridge.org/core/journals/behavioral-and-brain-sciences/article/generalization-similarity-and-bayesian-inference/595CAA321C9C56270C624057021DE77A">Generalization,
similarity, and Bayesian inference</a> - <strong><em>Behavioral and
Brain Sciences</em></strong>, 2001. [<a
href="https://scholar.google.com/scholar?cluster=14074987155133342565">All
Versions</a>]. [<a
href="http://web.mit.edu/cocosci/archive/Papers/tenenbaum_griffiths01.pdf">Preprint</a>].
Shepard has argued that a universal law should govern generalization
across different domains of perception and cognition, as well as across
organisms from different species or even different planets. Starting
with some basic assumptions about natural kinds, he derived an
exponential decay function as the form of the universal generalization
gradient, which accords strikingly well with a wide range of empirical
data. However, his original formulation applied only to the ideal case
of generalization from a single encountered stimulus to a single novel
stimulus, and for stimuli that can be represented as points in a
continuous metric psychological space. The authors recast Shepard’s
theory in a more general Bayesian framework and show how this naturally
extends his approach to the more realistic situation of generalizing
from multiple consequential stimuli with arbitrary representational
structure. This framework also subsumes a version of Tversky’s
set-theoretic model of similarity, which is conventionally thought of as
the primary alternative to Shepard’s continuous metric space model of
similarity and generalization.</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper_files/paper/1998/hash/d010396ca8abf6ead8cacc2c2f2f26c7-Abstract.html">Bayesian
modeling of human concept learning</a> -
<strong><em>NeurIPS’98</em></strong>, 1998. [<a
href="https://scholar.google.com/scholar?cluster=3772493362518191863">All
Versions</a>]. [<a
href="http://web.mit.edu/cocosci/archive/Papers/bayes.pdf">Preprint</a>].
This work considers the problem of learning concepts from small numbers
of positive examples, a feat which humans perform routinely but which
computers are rarely capable of. Bridging machine learning and cognitive
science perspectives, this work presents both theoretical analysis and
an empirical study with human subjects for the simple task oflearning
concepts corresponding to axis-aligned rectangles in a multidimensional
feature space. Existing learning models, when applied to this task,
cannot explain how subjects generalize from only a few examples of the
concept. The author proposes a principled Bayesian model based on the
assumption that the examples are a random sample from the concept to be
learned. The model gives precise fits to human behavior on this simple
task and provides qualitati ve insights into more complex, realistic
cases of concept learning.</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/1999/hash/86d7c8a08b4aaa1bc7c599473f5dddda-Abstract.html">Rules
and Similarity in Concept Learning</a> -
<strong><em>NeurIPS’99</em></strong>, 1999. [<a
href="https://scholar.google.com/scholar?cluster=10968021160883668417">All
Versions</a>]. [<a
href="http://web.mit.edu/cocosci/archive/Papers/nips99preprint.pdf">Preprint</a>].
This paper argues that two apparently distinct modes of generalizing
concepts - abstracting rules and computing similarity to exemplars -
should both be seen as special cases of a more general Bayesian learning
framework. Bayes explains the specific workings of these two modes -
which rules are abstracted, how similarity is measured - as well as why
generalization should appear rule- or similarity-based in different
situations. This analysis also suggests why the rules/similarity
distinction, even if not computationally fundamental, may still be
useful at the algorithmic level as part of a principled approximation to
fully Bayesian learning.</p></li>
<li><p><a
href="https://www.cell.com/AJHG/fulltext/S1364-6613(06)00134-3">Theory-based
Bayesian models of inductive learning and reasoning</a> -
<strong><em>Trends in Cognitive Sciences</em></strong>, 2006. [<a
href="https://scholar.google.com/scholar?cluster=6741344960992898446">All
Versions</a>]. [<a
href="http://www.charleskemp.com/papers/TenenbaumGK06.pdf">Preprint</a>].
Inductive inference allows humans to make powerful generalizations from
sparse data when learning about word meanings, unobserved properties,
causal relationships, and many other aspects of the world. Traditional
accounts of induction emphasize either the power of statistical
learning, or the importance of strong constraints from structured domain
knowledge, intuitive theories or schemas. This paper argues that both
components are necessary to explain the nature, use and acquisition of
human knowledge, and the authors introduce a theory-based Bayesian
framework for modeling inductive learning and reasoning as statistical
inferences over structured knowledge representations.</p></li>
<li><p><a
href="https://psycnet.apa.org/doiLanding?doi=10.1037%2F0033-295X.114.2.245">Word
learning as Bayesian inference</a> - <strong><em>Psychological
Review</em></strong>, 2007. [<a
href="https://scholar.google.com/scholar?cluster=5476233692839102256">All
Versions</a>]. [<a
href="https://tallinzen.net/media/readings/xu_tenenbaum_2007.pdf">Preprint</a>].
The authors present a Bayesian framework for understanding how adults
and children learn the meanings of words. The theory explains how
learners can generalize meaningfully from just one or a few positive
examples of a novel word’s referents, by making rational inductive
inferences that integrate prior knowledge about plausible word meanings
with the statistical structure of the observed examples. The theory
addresses shortcomings of the two best known approaches to modeling word
learning, based on deductive hypothesis elimination and associative
learning. Three experiments with adults and children test the Bayesian
account’s predictions in the context of learning words for object
categories at multiple levels of a taxonomic hierarchy. Results provide
strong support for the Bayesian account over competing accounts, in
terms of both quantitative model fits and the ability to explain
important qualitative phenomena. Several extensions of the basic theory
are discussed, illustrating the broader potential for Bayesian models of
word learning.</p></li>
<li><p><a
href="https://www.science.org/doi/full/10.1126/science.1192788">How to
Grow a Mind: Statistics, Structure, and Abstraction</a> -
<strong><em>Science</em></strong>, 2011. [<a
href="https://scholar.google.com/scholar?cluster=2667398573353002097">All
Versions</a>]. [<a
href="https://cocosci.princeton.edu/tom/papers/growamind.pdf">Preprint</a>].
This review describes recent approaches to reverse-engineering human
learning and cognitive development and, in parallel, engineering more
humanlike machine learning systems. Computational models that perform
probabilistic inference over hierarchies of flexibly structured
representations can address some of the deepest questions about the
nature and origins of human thought: How does abstract knowledge guide
learning and reasoning from sparse data? What forms does our knowledge
take, across different domains and tasks? And how is that abstract
knowledge itself acquired?</p></li>
<li><p><a
href="https://www.science.org/doi/full/10.1126/science.aab3050">Human-level
concept learning through probabilistic program induction</a> -
<strong><em>Science</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=11844685101409624506">All
Versions</a>]. [<a
href="https://ai6034.mit.edu/wiki/images/LakeDec2015.pdf">Preprint</a>].
[<a
href="https://cims.nyu.edu/~brenden/LakeEtAl2015Science_supp.pdf">Supplementary
Material</a>]. People learning new concepts can often generalize
successfully from just a single example, yet machine learning algorithms
typically require tens or hundreds of examples to perform with similar
accuracy. People can also use learned concepts in richer ways than
conventional algorithms—for action, imagination, and explanation. This
work presents a computational model that captures these human learning
abilities for a large class of simple visual concepts: handwritten
characters from the world’s alphabets. The model represents concepts as
simple programs that best explain observed examples under a Bayesian
criterion. On a challenging one-shot classification task, the model
achieves human-level performance while outperforming recent deep
learning approaches.</p></li>
<li><p><a
href="https://www.cambridge.org/core/journals/behavioral-and-brain-sciences/article/building-machines-that-learn-and-think-like-people/A9535B1D745A0377E16C590E14B94993">Building
Machines That Learn and Think Like People</a> - <strong><em>Behavioral
and Brain Sciences</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=8504723689348856287">All
Versions</a>]. [<a
href="https://leylaroksancaglar.github.io/Caglar_Hanson_2017.pdf">Preprint</a>].
Recent progress in artificial intelligence has renewed interest in
building systems that learn and think like people. Many advances have
come from using deep neural networks trained end-to-end in tasks such as
object recognition, video games, and board games, achieving performance
that equals or even beats that of humans in some respects. Despite their
biological inspiration and performance achievements, these systems
differ from human intelligence in crucial ways. The authors review
progress in cognitive science suggesting that truly human-like learning
and thinking machines will have to reach beyond current engineering
trends in both what they learn and how they learn it. Specifically, the
authors argue that these machines should (1) build causal models of the
world that support explanation and understanding, rather than merely
solving pattern recognition problems; (2) ground learning in intuitive
theories of physics and psychology to support and enrich the knowledge
that is learned; and (3) harness compositionality and learning-to-learn
to rapidly acquire and generalize knowledge to new tasks and situations.
The authors suggest concrete challenges and promising routes toward
these goals that can combine the strengths of recent neural network
advances with more structured cognitive models.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41562-024-01991-9">Building
machines that learn and think with people</a> - <strong><em>Nature Human
Behavior</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=4420595706578245444">All
Versions</a>]. [<a
href="https://arxiv.org/abs/2408.03943">Preprint</a>]. This perspective
shows how the science of collaborative cognition can be put to work to
engineer systems that really can be called ‘thought partners’, systems
built to meet humans’ expectations and complement humans’ limitations.
The authors lay out several modes of collaborative thought in which
humans and artificial intelligence thought partners can engage, and they
propose desiderata for human-compatible thought partnerships. Drawing on
motifs from computational cognitive science, this work motivates an
alternative scaling path for the design of thought partners and
ecosystems around their use through a Bayesian lens, whereby the
constructed partners actively build and reason over models of the human
and world.</p></li>
<li><p><a
href="http://web.mit.edu/cocosci/archive/Papers/cogsci01_final.pdf">The
rational basis of representativeness</a> -
<strong><em>CogSci’01</em></strong>, 2001. [<a
href="https://scholar.google.com/scholar?cluster=11464039134248091466&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2011/hash/2c89109d42178de8a367c0228f169bf8-Abstract.html">Testing
a Bayesian Measure of Representativeness Using a Large Image
Database</a> - <strong><em>NeurIPS’11</em></strong>, 2011. [<a
href="https://scholar.google.com/scholar?cluster=8576570792794301292&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://cocosci.princeton.edu/tom/papers/abbott_cogsci2012_wordnet.pdf">Constructing
a hypothesis space from the Web for large-scale Bayesian word
learning</a> - <strong><em>CogSci’12</em></strong>, 2012. [<a
href="https://scholar.google.com/scholar?cluster=9266416266046851766&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://escholarship.org/content/qt1md755ng/qt1md755ng.pdf">Modeling
rules and similarity in colexification</a> -
<strong><em>CogSci’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=11578380234126546169">All
Versions</a>]. Rule- and similarity-based generalization in
colexification.</p></li>
<li><p><a
href="https://www.science.org/doi/full/10.1126/sciadv.adg2488">Human-level
few-shot concept induction through minimax entropy learning</a> -
<strong><em>Science Advances</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?&amp;cluster=9084477652494351940">All
Versions</a>]. This paper introduces a computational model designed to
emulate human inductive reasoning on abstract reasoning tasks, such as
those in IQ tests, using a minimax entropy approach. This method
combines identifying the most effective constraints on data via minimum
entropy with determining the best combination of them via maximum
entropy.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="generative-model">Generative Model</h4>
<ul>
<li><p><a
href="https://github.com/YuzheSHI/generative-modeling-explained">Generative
Modeling Explained</a> - <strong><em>Statistical Machine Learning
Tutorials</em></strong>, 2022. This tutorial on generative modeling is
in part of Statistical Machine Learning Tutorial by Ying Nian Wu at UCLA
Statistics. The tutorial goes over the key equations and algorithms for
learning recent generative models, including energy-based models,
diffusion/score-based models, autoregressive/flow-based models, VAEs,
and GANs, and explains the connections between these models.</p></li>
<li><p><a
href="https://www.taylorfrancis.com/books/mono/10.1201/9780429258411/bayesian-data-analysis-andrew-gelman-donald-rubin-john-carlin-hal-stern">Bayesian
Data Analysis</a> - <strong><em>Chapman and Hall/CRC</em></strong>,
1995. [<a
href="https://scholar.google.com/scholar?cluster=5067275302121330689&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Don Rubin’s introductory book on Bayesian
models.</p></li>
<li><p><a
href="https://link.springer.com/article/10.1023/A:1007925832420">Filters,
random fields and maximum entropy (FRAME): Towards a unified theory for
texture modeling</a> - <strong><em>International Journal of Computer
Vision</em></strong>, 1998. [<a
href="https://scholar.google.com/scholar?cluster=11604954524863138240">All
Versions</a>]. [<a
href="https://dash.harvard.edu/bitstream/handle/1/3637117/Mumford_FRAME.pdf?sequence=1">Preprint</a>].
This article presents a statistical theory for texture modeling. This
theory combines filtering theory and Markov random field modeling
through the maximum entropy principle, and interprets and clarifies many
previous concepts and methods for texture analysis and synthesis from a
unified point of view. The theory characterizes the ensemble of images I
with the same texture appearance by a probability distribution f(I) on a
random field, and the objective of texture modeling is to make inference
about f(I), given a set of observed texture examples.</p></li>
<li><p><a
href="https://www.annualreviews.org/content/journals/10.1146/annurev.psych.55.090902.142005">Object
Perception as Bayesian Inference</a> - <strong><em>Annual Review of
Psychology</em></strong>, 2004. [<a
href="https://scholar.google.com/scholar?cluster=1611451804975333652">All
Versions</a>]. [<a
href="https://www.cs.jhu.edu/~ayuille/pubs/ucla/A189_dkersten_ARP2004.pdf">Preprint</a>].
We perceive the shapes and material properties of objects quickly and
reliably despite the complexity and objective ambiguities of natural
images. Typical images are highly complex because they consist of many
objects embedded in background clutter. Moreover, the image features of
an object are extremely variable and ambiguous owing to the effects of
projection, occlusion, background clutter, and illumination. The very
success of everyday vision implies neural mechanisms, yet to be
understood, that discount irrelevant information and organize ambiguous
or noisy local image features into objects and surfaces. Recent work in
Bayesian theories of visual perception has shown how complexity may be
managed and ambiguity resolved through the task-dependent, probabilistic
integration of prior object knowledge with image features.</p></li>
<li><p><a
href="https://www.ams.org/journals/qam/2019-77-02/S0033-569X-2018-01528-5/home.html">A
tale of three probabilistic families: Discriminative, descriptive, and
generative models</a> - <strong><em>Quarterly of Applied
Mathematics</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=6129609629126793774">All
Versions</a>]. [<a
href="http://www.stat.ucla.edu/~ywu/QAM2018.pdf">Preprint</a>]. The
pattern theory of Grenander is a mathematical framework where patterns
are represented by probability models on random variables of algebraic
structures. In this paper, the authors review three families of
probability models, namely, the discriminative models, the descriptive
models, and the generative models. A discriminative model is in the form
of a classifier. It specifies the conditional probability of the class
label given the input signal. A descriptive model specifies the
probability distribution of the signal, based on an energy function
defined on the signal. A generative model assumes that the signal is
generated by some latent variables via a transformation. The authors
shall review these models within a common framework and explore their
connections, and shall also review the recent developments that take
advantage of the high approximation capacities of deep neural
networks.</p></li>
<li><p><a href="https://www.jstor.org/stable/43638808?seq=1">From
information scaling of natural images to regimes of statistical
models</a> - <strong><em>Quarterly of Applied Mathematics</em></strong>,
2008. [<a
href="https://scholar.google.com/scholar?cluster=17387130978932998303">All
Versions</a>]. [<a
href="http://www.stat.ucla.edu/~sczhu/papers/Quarterly_final.pdf">Preprint</a>].
One fundamental property of natural image data that distinguishes vision
from other sensory tasks such as speech recognition is that scale plays
a profound role in image formation and interpretation. Specifically,
visual objects can appear at a wide range of scales in the images due to
the change of viewing distance as well as camera resolution. The same
objects appearing at different scales produce different image data with
different statistical properties. In particular, this work shows that
the entropy rate of the image data changes over scale. Moreover, the
inferential uncertainty changes over scale too. The authors call these
changes information scaling. They then examine both empirically and
theoretically two prominent and yet largely isolated classes of image
models, namely, wavelet sparse coding models and Markov random field
models. The results indicate that the two classes of models are
appropriate for two different entropy regimes: sparse coding targets low
entropy regimes, whereas Markov random fields are appropriate for high
entropy regimes. Because information scaling connects different entropy
regimes, both sparse coding and Markov random fields are necessary for
representing natural image data, and information scaling triggers
transitions between these two regimes.</p></li>
<li><p><a href="https://proceedings.mlr.press/v48/xiec16.html">A Theory
of Generative ConvNet</a> - <strong><em>ICML’16</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=11062907630625111054">All
Versions</a>]. The authors show that a generative random field model,
which they call generative ConvNet, can be derived from the commonly
used discriminative ConvNet, by assuming a ConvNet for multi-category
classification and assuming one of the category is a base category
generated by a reference distribution. For a further assumption that the
non-linearity in the ConvNet is Rectified Linear Unit (ReLU) and the
reference distribution is Gaussian white noise, then a generative
ConvNet model that is unique among energy-based models is obtained: The
model is piecewise Gaussian, and the means of the Gaussian pieces are
defined by an auto-encoder, where the filters in the bottom-up encoding
become the basis functions in the top-down decoding, and the binary
activation variables detected by the filters in the bottom-up
convolution process become the coefficients of the basis functions in
the top-down deconvolution process.</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/abstract/document/8519332">Cooperative
Training of Descriptor and Generator Networks</a> - <strong><em>IEEE
Transactions on Pattern Analysis and Machine Intelligence</em></strong>,
2018. [<a
href="https://scholar.google.com/scholar?cluster=18202808849093155435">All
Versions</a>]. This paper studies the cooperative training of two
generative models for image modeling and synthesis. Both models are
parametrized by convolutional neural networks (ConvNets). The first
model is a deep energy-based model, whose energy function is defined by
a bottom-up ConvNet, which maps the observed image to the energy. We
call it the descriptor network. The second model is a generator network,
which is a non-linear version of factor analysis. It is defined by a
top-down ConvNet, which maps the latent factors to the observed image.
The maximum likelihood learning algorithms of both models involve MCMC
sampling such as Langevin dynamics. This work observes that the two
learning algorithms can be seamlessly interwoven into a cooperative
learning algorithm that can train both models simultaneously.
Specifically, within each iteration of the cooperative learning
algorithm, the generator model generates initial synthesized examples to
initialize a finite-step MCMC that samples and trains the energy-based
descriptor model. After that, the generator model learns from how the
MCMC changes its synthesized examples. That is, the descriptor model
teaches the generator model by MCMC, so that the generator model
accumulates the MCMC transitions and reproduces them by direct ancestral
sampling.</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2020/hash/fa3060edb66e6ff4507886f9912e1ab9-Abstract.html">Learning
Latent Space Energy-Based Prior Model</a> -
<strong><em>NeurIPS’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=9945264852135249894">All
Versions</a>]. [<a
href="https://bpucla.github.io/latent-space-ebm-prior-project/">Project</a>].
[<a href="https://github.com/bpucla/latent-space-EBM-prior">Code</a>]. A
milestone paper on Latent Energy-Based Model.</p></li>
<li><p><a href="https://openreview.net/forum?id=v_1Soh8QUNc">Learning
Energy-Based Models by Diffusion Recovery Likelihood</a> -
<strong><em>ICLR’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=4399294843209736764">All
Versions</a>]. [<a
href="https://github.com/ruiqigao/recovery_likelihood">Code</a>].</p></li>
<li><p><a
href="https://openreview.net/forum?id=PxTIG12RRHS&amp;utm_campaign=NLP%20News&amp;utm_medium=email&amp;utm_source=Revue%20newsletter">Score-Based
Generative Modeling through Stochastic Differential Equations</a> -
<strong><em>ICLR’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=14592788616550656262">All
Versions</a>].</p></li>
<li><p><a href="http://proceedings.mlr.press/v119/li20i.html">Latent
Space Factorisation and Manipulation via Matrix Subspace Projection</a>
- <strong><em>ICML’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=9592355331559392684">All
Versions</a>].</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/abstract/document/6796444">Minimax
entropy principle and its application to texture modeling</a> -
<strong><em>Neural Computing</em></strong>, 1997. [<a
href="https://scholar.google.com/scholar?cluster=407872717119429940">All
Versions</a>]. [<a
href="https://www.dam.brown.edu/people/mumford/vision/papers/1997e--MinimaxEntropy-NC.pdf">Preprint</a>].
This article proposes a general theory and methodology, called the
minimax entropy principle, for building statistical models for images
(or signals) in a variety of applications. This principle consists of
two parts. The first is the maximum entropy principle for feature
binding (or fusion): for a given set of observed feature statistics, a
distribution can be built to bind these feature statistics together by
maximizing the entropy over all distributions that reproduce them. The
second part is the minimum entropy principle for feature selection:
among all plausible sets of feature statistics, we choose the set whose
maximum entropy distribution has the minimum entropy. Computational and
inferential issues in both parts are addressed; in particular, a feature
pursuit procedure is proposed for approximately selecting the optimal
set of features. The minimax entropy principle is then corrected by
considering the sample variation in the observed feature statistics, and
an information criterion for feature pursuit is derived. The minimax
entropy principle is applied to texture modeling, where a novel Markov
random field (MRF) model, called FRAME (filter, random field, and
minimax entropy), is derived, and encouraging results are obtained in
experiments on a variety of texture images.</p></li>
<li><p><a
href="https://www.tandfonline.com/doi/abs/10.1080/01621459.1999.10473879">Parameter
Expansion for Data Augmentation</a> - <strong><em>Journal of the
American Statistical Association</em></strong>, 1999. [<a
href="https://scholar.google.com/scholar?cluster=15342818142955984734">All
Versions</a>]. [<a
href="http://www.stat.ucla.edu/~ywu/research/papers/PXDA.pdf">Preprint</a>].
Viewing the observed data of a statistical model as incomplete and
augmenting its missing parts are useful for clarifying concepts and
central to the invention of two well-known statistical algorithms:
expectation-maximization (EM) and data augmentation. Recently, the
authors demonstrated that expanding the parameter space along with
augmenting the missing data is useful for accelerating iterative
computation in an EM algorithm. The main purpose of this article is to
rigorously define a parameter expanded data augmentation (PX-DA)
algorithm and to study its theoretical properties. The PX-DA is a
special way of using auxiliary variables to accelerate Gibbs sampling
algorithms and is closely related to reparameterization
techniques.</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/abstract/document/1000239">Image
segmentation by data-driven markov chain monte carlo</a> -
<strong><em>IEEE Transactions on Pattern Analysis and Machine
Intelligence</em></strong>, 2002. [<a
href="https://scholar.google.com/scholar?cluster=3461400072144667491">All
Versions</a>]. [<a
href="http://www.stat.ucla.edu/~sczhu/papers/DDMCMC_reprint.pdf">Preprint</a>].
This paper presents a computational paradigm called Data-Driven Markov
Chain Monte Carlo (DDMCMC) for image segmentation in the Bayesian
statistical framework. The paper contributes to image segmentation in
four aspects. First, it designs efficient and well-balanced Markov Chain
dynamics to explore the complex solution space and, thus, achieves a
nearly global optimal solution independent of initial segmentations.
Second, it presents a mathematical principle and a K-adventurers
algorithm for computing multiple distinct solutions from the Markov
chain sequence and, thus, it incorporates intrinsic ambiguities in image
segmentation. Third, it utilizes data-driven (bottom-up) techniques,
such as clustering and edge detection, to compute importance proposal
probabilities, which drive the Markov chain dynamics and achieve
tremendous speedup in comparison to the traditional jump-diffusion
methods. Fourth, the DDMCMC paradigm provides a unifying framework in
which the role of many existing segmentation algorithms, such as, edge
detection, clustering, region growing, split-merge, snake/balloon, and
region competition, are revealed as either realizing Markov chain
dynamics or computing importance proposal probabilities. Thus, the
DDMCMC paradigm combines and generalizes these segmentation methods in a
principled way.</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2006/file/87f4d79e36d68c3031ccf6c55e9bbd39-Paper.pdf">Efficient
Learning of Sparse Representations with an Energy-Based Model</a> -
<strong><em>NeurIPS’06</em></strong>, 2006. [<a
href="https://scholar.google.com/scholar?cluster=2247668190782691760">All
Versions</a>].</p></li>
<li><p><a href="http://yann.lecun.com/exdb/publis/orig/lecun-06.pdf">A
Tutorial on Energy-Based Learning</a> - <strong><em>Predicting
Structured Data, MIT Press</em></strong>, 2006. [<a
href="https://scholar.google.com/scholar?cluster=8819502341081664768&amp;hl=en&amp;as_sdt=0,5">All
Versiosn</a>]. Yann LeCun’s tutorial on energy-based learning.</p></li>
<li><p><a href="https://arxiv.org/abs/1511.06434">Unsupervised
Representaton Learning with Deep Convolutional Generative Adversarial
Networks</a> - <strong><em>ICLR’16</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=3321343160055675528&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://www.jmlr.org/papers/v20/18-173.html">Analysis of
Langevin Monte Carlo via Convex Optimization</a> - <strong><em>Journal
of Machine Learning Research</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=5305860199396047317">All
Versions</a>]. This paper provides new insights on the Unadjusted
Langevin Algorithm. The authors show that this method can be formulated
as the first order optimization algorithm for an objective functional
defined on the Wasserstein space of order <span
class="math inline">2</span>. Using this interpretation and techniques
borrowed from convex optimization, the authors give a non-asymptotic
analysis of this method to sample from log-concave smooth target
distribution on <span class="math inline">ℝ<sup><em>d</em></sup></span>.
Based on this interpretation, the authors propose two new methods for
sampling from a non-smooth target distribution. These new algorithms are
natural extensions of the Stochastic Gradient Langevin Dynamics (SGLD)
algorithm, which is a popular extension of the Unadjusted Langevin
Algorithm for largescale Bayesian inference. Using the optimization
perspective, the authors provide non-asymptotic convergence analysis for
the newly proposed methods.</p></li>
<li><p><a
href="https://www.science.org/doi/full/10.1126/science.aag2612">A
generative vision model that trains with high data efficiency and breaks
text-based CAPTCHAs</a> - <strong><em>Science</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=1478382321633671444">All
Versions</a>]. [<a
href="https://www.cs.jhu.edu/~ayuille/JHUcourses/ProbabilisticModelsOfVisualCognition2020/Lec22/GeorgeCAPCHAS.pdf">Preprint</a>].
Learning from a few examples and generalizing to markedly different
situations are capabilities of human visual intelligence that are yet to
be matched by leading machine learning models. By drawing inspiration
from systems neuroscience, this work introduces a probabilistic
generative model for vision in which message-passing–based inference
handles recognition, segmentation, and reasoning in a unified way. The
model demonstrates excellent generalization and occlusion-reasoning
capabilities and outperforms deep neural networks on a challenging scene
text recognition benchmark while being 300-fold more data efficient. In
addition, the model fundamentally breaks the defense of modern
text-based CAPTCHAs (Completely Automated Public Turing test to tell
Computers and Humans Apart) by generatively segmenting characters
without CAPTCHA-specific heuristics.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0010028516302766">Where
do hypotheses come from?</a> - <strong><em>Cognitive
Psychology</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=17480320046655923235">All
Versions</a>]. [<a
href="https://gershmanlab.com/pubs/Dasgupta17.pdf">Preprint</a>]. Why
are human inferences sometimes remarkably close to the Bayesian ideal
and other times systematically biased? In particular, why do humans make
near-rational inferences in some natural domains where the candidate
hypotheses are explicitly available, whereas tasks in similar domains
requiring the self-generation of hypotheses produce systematic
deviations from rational inference. This work proposes that these
deviations arise from algorithmic processes approximating Bayes’ rule.
Specifically in our account, hypotheses are generated stochastically
from a sampling process, such that the sampled hypotheses form a Monte
Carlo approximation of the posterior.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="nonparametric-model">Nonparametric Model</h4>
<ul>
<li><p><a href="https://www.jstor.org/stable/2958008?seq=1">A Bayesian
Analysis of Some Non-parametric Problems</a> - <strong><em>The Annals of
Statistics</em></strong>, 1973. [<a
href="https://scholar.google.com/scholar?cluster=3969163427460060902">All
Versions</a>]. [<a
href="https://people.stat.sc.edu/hansont/stat740/Ferguson1973.pdf">Preprint</a>].
A classic review on non-parametric problems.</p></li>
<li><p><a
href="https://people.eecs.berkeley.edu/~jordan/courses/281B-spring04/readings/antoniak.pdf">Mixtures
of Dirichlet Process with Applications to Bayesian Nonparametric
Problems</a> - <strong><em>The Annals of Statistics</em></strong>, 1974.
[<a
href="https://scholar.google.com/scholar?cluster=17937202534282344046&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper on Dirichlet Process modeling for
non-parametric problems.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0022000000917112">Latent
Semantic Indexing: A Probabilistic Analysis</a> - <strong><em>Journal of
Computer and System Sciences</em></strong>, 2000. [<a
href="https://scholar.google.com/scholar?cluster=7296120469860429813&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper on hierarchical topic model.</p></li>
<li><p><a
href="https://projecteuclid.org/journals/statistical-science/volume-19/issue-1/Nonparametric-Bayesian-Data-Analysis/10.1214/088342304000000017.full">Nonparametric
Bayesian Data Analysis</a> - <strong><em>Statistical
Science</em></strong>, 2004. [<a
href="https://scholar.google.com/scholar?cluster=13476170780072319995">All
Versions</a>]. This paper reviews the current state of nonparametric
Bayesian inference. The discussion follows a list of important
statistical inference problems, including density estimation,
regression, survival analysis, hierarchical models and model validation.
For each inference problem the authors review relevant nonparametric
Bayesian models and approaches including Dirichlet process (DP) models
and variations, Pólya trees, wavelet based models, neural network
models, spline regression, CART, dependent DP models and model
validation with DP and Pólya tree extensions of parametric
models.</p></li>
<li><p><a
href="https://www.pnas.org/doi/abs/10.1073/pnas.0307752101">Finding
scientific topics</a> - <strong><em>Proceedings of the National Academy
of Sciences</em></strong>, 2004. [<a
href="https://scholar.google.com/scholar?cluster=17382767110929995134">All
Versions</a>]. A first step in identifying the content of a document is
determining which topics that document addresses. This paper describes a
generative model for documents, in which each document is generated by
choosing a distribution over topics and then choosing each word in the
document from a topic selected according to this distribution. The
authors then present a Markov chain Monte Carlo algorithm for inference
in this model. The authors use this algorithm to analyze abstracts from
PNAS by using Bayesian model selection to establish the number of
topics. This work shows that the extracted topics capture meaningful
structure in the data, consistent with the class designations provided
by the authors of the articles, and outline further applications of this
analysis, including identifying “hot topics” by examining temporal
dynamics and tagging abstracts to illustrate semantic content.</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2003/file/7b41bfa5085806dfa24b8c9de0ce567f-Paper.pdf">Hierarchical
topic models and the nested Chinese restaurant process</a> -
<strong><em>NeurIPS’03</em></strong>, 2003. [<a
href="https://scholar.google.com/scholar?cluster=15040818675282958700&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper for nested Chinese restaurant
process.</p></li>
<li><p><a
href="https://www.aaai.org/Papers/AAAI/2006/AAAI06-061.pdf">Learning
Systems of Concepts with an Infinite Relational Model</a> -
<strong><em>AAAI’06</em></strong>, 2006. [<a
href="https://scholar.google.com/scholar?cluster=3207350432755252565&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://dl.acm.org/doi/abs/10.1145/1667053.1667056">The
nested chinese restaurant process and bayesian nonparametric inference
of topic hierarchies</a> - <strong><em>Journal of the ACM</em></strong>,
2010. [<a
href="https://scholar.google.com/scholar?cluster=8216933258869737505&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://mlg.eng.cam.ac.uk/zoubin/papers/ibptr.pdf">Infinite Latent
Feature Models and the Indian Buffet Process</a> - <strong><em>Gatsby
Computational Neuroscience Unit Technical Report 2005-001</em></strong>,
2005. [<a
href="https://scholar.google.com/scholar?cluster=13180738480564152907&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://jmlr.org/papers/v12/griffiths11a.html">The
Indian Buffet Process: An Introduction and Review</a> -
<strong><em>Journal of Machine Learning Research</em></strong>, 2011.
[<a
href="https://scholar.google.com/scholar?cluster=6301314251995890943">All
Versions</a>]. The Indian buffet process is a stochastic process
defining a probability distribution over equivalence classes of sparse
binary matrices with a finite number of rows and an unbounded number of
columns. This distribution is suitable for use as a prior in
probabilistic models that represent objects using a potentially infinite
array of features, or that involve bipartite graphs in which the size of
at least one class of nodes is unknown. This work gives a detailed
derivation of this distribution, and illustrate its use as a prior in an
infinite latent feature model. The authors then review recent
applications of the Indian buffet process in machine learning, discuss
its extensions, and summarize its connections to other stochastic
processes.</p></li>
<li><p><a
href="https://dl.acm.org/doi/abs/10.5555/3020336.3020347">Nonparametric
Bayesian Logic</a> - <strong><em>UAI’05</em></strong>, 2005. [<a
href="https://scholar.google.com/scholar?cluster=18267211625980322095">All
Versions</a>]. [<a
href="https://www.cs.ubc.ca/~nando/papers/npblog.pdf">Preprint</a>]. The
Bayesian Logic (BLOG) language was recently developed for defining
first-order probability models over worlds with unknown numbers of
objects. It handles important problems in AI, including data association
and population estimation. This paper extends BLOG by adopting
generative processes over function spaces — known as nonparametrics in
the Bayesian literature. This work introduces syntax for reasoning about
arbitrary collections of objects, and their properties, in an intuitive
manner. By exploiting exchangeability, distributions over unknown
objects and their attributes are cast as Dirichlet processes, which
resolve difficulties in model selection and inference caused by varying
numbers of objects.</p></li>
<li><p><a
href="https://dl.acm.org/doi/abs/10.5555/3020419.3020485">Infinite
Hidden Relational Models</a> - <strong><em>UAI’06</em></strong>, 2006.
[<a
href="https://scholar.google.com/scholar?cluster=2143172296528388141">All
Versions</a>]. [<a
href="https://www.dbs.ifi.lmu.de/~yu_k/uai06_relation.pdf">Preprint</a>].
Relational learning analyzes the probabilistic constraints between the
attributes of entities and relationships. This work extends the
expressiveness of relational models by introducing for each entity (or
object) an infinite-dimensional latent variable as part of a Dirichlet
process (DP) mixture model. This work discusses inference in the model,
which is based on a DP Gibbs sampler, i.e., the Chinese restaurant
process. The authors extended the Chinese restaurant process to be
applicable to relational modeling.</p></li>
<li><p><a
href="https://alchemy.cs.washington.edu/papers/kok07/kok07.pdf">Statistical
Predicate Invention</a> - <strong><em>ICML’07</em></strong>, 2007. [<a
href="https://scholar.google.com/scholar?cluster=17009312281859401704">All
Versions</a>]. This work proposes statistical predicate invention as a
key problem for statistical relational learning. SPI is the problem of
discovering new concepts, properties and relations in structured data,
and generalizes hidden variable discovery in statistical models and
predicate invention in ILP. This work proposes an initial model for SPI
based on second-order Markov logic, in which predicates as well as
arguments can be variables, and the domain of discourse is not fully
known in advance. The proposed approach iteratively refines clusters of
symbols based on the clusters of symbols they appear in atoms with
(e.g., it clusters relations by the clusters of the objects they
relate).</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="bayesian-optimization">Bayesian Optimization</h4>
<ul>
<li><p><a
href="https://ieeexplore.ieee.org/abstract/document/7352306">Taking the
Human Out of the Loop: A Review of Bayesian Optimization</a> -
<strong><em>Proceedings of the IEEE</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=2039456143890648437">All
Versions</a>]. [<a
href="https://www.cs.princeton.edu/~rpa/pubs/shahriari2016loop.pdf">Preprint</a>].
Big Data applications are typically associated with systems involving
large numbers of users, massive complex software systems, and
large-scale heterogeneous computing and storage architectures. The
construction of such systems involves many distributed design choices.
The end products (e.g., recommendation systems, medical analysis tools,
real-time game engines, speech recognizers) thus involve many tunable
configuration parameters. These parameters are often specified and
hard-coded into the software by various developers or teams. If
optimized jointly, these parameters can result in significant
improvements. Bayesian optimization is a powerful tool for the joint
optimization of design choices that is gaining great popularity in
recent years. It promises greater automation so as to increase both
product quality and human productivity. This review paper introduces
Bayesian optimization, highlights some of its methodological aspects,
and showcases a wide range of applications.</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2012/hash/05311655a15b75fab86956663e1819cd-Abstract.html">Practical
Bayesian Optimization of Machine Learning Algorithms</a> -
<strong><em>NeurIPS’12</em></strong>, 2012. [<a
href="https://scholar.google.com/scholar?cluster=14442949298925775705">All
Versions</a>]. The use of machine learning algorithms frequently
involves careful tuning of learning parameters and model
hyperparameters. Unfortunately, this tuning is often a “black art”
requiring expert experience, rules of thumb, or sometimes brute-force
search. There is therefore great appeal for automatic approaches that
can optimize the performance of any given learning algorithm to the
problem at hand. This work considers this problem through the framework
of Bayesian optimization, in which a learning algorithm’s generalization
performance is modeled as a sample from a Gaussian process (GP). The
authors show that certain choices for the nature of the GP, such as the
type of kernel and the treatment of its hyperparameters, can play a
crucial role in obtaining a good optimizer that can achieve expert-level
performance. The authors describe new algorithms that take into account
the variable cost (duration) of learning algorithm experiments and that
can leverage the presence of multiple cores for parallel
experimentation. These proposed algorithms improve on previous automatic
procedures and can reach or surpass human expert-level optimization for
many algorithms including Latent Dirichlet Allocation, Structured SVMs
and convolutional neural networks.</p></li>
<li><p><a href="https://arxiv.org/abs/1807.02811">A Tutorial on Bayesian
Optimization</a> - 2018. [<a
href="https://scholar.google.com/scholar?cluster=7971934771645047583">All
Versions</a>]. Bayesian optimization is an approach to optimizing
objective functions that take a long time (minutes or hours) to
evaluate. It is best-suited for optimization over continuous domains of
less than 20 dimensions, and tolerates stochastic noise in function
evaluations. It builds a surrogate for the objective and quantifies the
uncertainty in that surrogate using a Bayesian machine learning
technique, Gaussian process regression, and then uses an acquisition
function defined from this surrogate to decide where to sample. This
tutorial describes how Bayesian optimization works, including Gaussian
process regression and three common acquisition functions: expected
improvement, entropy search, and knowledge gradient. The authors then
discuss more advanced techniques, including running multiple function
evaluations in parallel, multi-fidelity and multi-information source
optimization, expensive-to-evaluate constraints, random environmental
conditions, multi-task Bayesian optimization, and the inclusion of
derivative information. The authors conclude with a discussion of
Bayesian optimization software and future research directions in the
field. This tutorial provides a generalization of expected improvement
to noisy evaluations, beyond the noise-free setting where it is more
commonly applied. This generalization is justified by a formal
decision-theoretic argument, standing in contrast to previous ad hoc
modifications.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="concepts">Concepts</h3>
<h4 id="theory-of-concepts">Theory of Concepts</h4>
<ul>
<li><p><a
href="https://plato.stanford.edu/entries/concepts/">Concepts</a> -
<strong><em>Plato Stanford</em></strong>. A collection of the
computational philosophical debates about the concepts.</p></li>
<li><p><a
href="https://en.wikipedia.org/wiki/Theory-theory">Theory-theory</a> -
<strong><em>Wikipedia</em></strong>. Wikipedia for the Theory theory, a
perspective that contextualizes concepts in theoretical (or empirical)
systems.</p></li>
<li><p><a href="https://hk1lib.org/book/3659332/11fa44">Conceptual
Change in Childhood</a> - <strong><em>MIT Press</em></strong>, 1985. [<a
href="https://scholar.google.com/scholar?cluster=11720022076481483465">All
Versions</a>]. Susan Carey’s book on the theory theory of concepts in
child development.</p></li>
<li><p><a
href="http://library.lol/main/6A8215E9BAEB77F198C98CD75C517E02">Words,
thoughts, and theories</a> - <strong><em>MIT Press</em></strong>, 1997.
[<a
href="https://scholar.google.com/scholar?cluster=16726462136203686735&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Alison Gopnik’s book that articulates and defends the
“theory theory” of cognitive and semantic development, the idea that
infants and young children, like scientists, learn about the world by
forming and revising theories-a view of the origins of knowledge and
meaning that has broad implications for cognitive science.</p></li>
<li><p><a href="https://psycnet.apa.org/record/1994-97940-009">The
Theory Theory</a> - <strong><em>Mapping the mind: Domain specificity in
cognition and culture, Cambridge University Press</em></strong>, 1994.
[<a
href="https://scholar.google.com/scholar?cluster=9397889700764191662&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Alison Gopnik’s original paper on the theory
theory.</p></li>
<li><p><a
href="https://hk1lib.org/book/844457/42178f?id=844457&amp;secret=42178f">The
Origin of Concepts</a> - <strong><em>Oxford University
Press</em></strong>, 2009. [<a
href="https://scholar.google.com/scholar?cluster=11493102398422813821&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Susan Carey’s extended book on the theory theory of
concepts in child development.</p></li>
<li><p><a href="https://osf.io/preprints/psyarxiv/xrnb2">What we mean
when we say semantic: A Consensus statement on the nomenclature of
semantic memory</a> - 2023. [<a
href="https://scholar.google.com/scholar?cluster=7464626532716945232">All
Versions</a>]. The aim of this multidisciplinary workgroup was to
establish consensus definitions for some of the major recurring
constructs in semantic research (e.g., concept, amodal, abstract). These
efforts yielded a glossary consisting of succinct definitions,
agreement, subjective confidence ratings, relevant theoretical
background, and principled dissenting views. These core definitions will
potentially yield benchmarks for aligning perspectives and improving
cross-disciplinary communication in semantic research.</p></li>
<li><p><a
href="https://psycnet.apa.org/record/2012-12791-001">Reconstructing
constructivism: Causal models, Bayesian learning mechanisms, and the
theory theory</a> - <strong><em>Psychological Bulletin</em></strong>,
2012. [<a
href="https://scholar.google.com/scholar?cluster=11218217347365817167&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Alison Gopnik’s review on the constructivism idea of
developmental research, including the theory theory of
concepts.</p></li>
<li><p><a
href="https://groups.psych.northwestern.edu/gentner/newpdfpapers/MedinGoldstoneGentner90.pdf">Similarity
involving attributes and relations: Judgments of similarity and
difference are not inverses</a> - <strong><em>Psychological
Science</em></strong>, 1990. [<a
href="https://scholar.google.com/scholar?cluster=13205938250772079784">All
Versions</a>]. Theory on similarity judgement by attributes and
relations.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="human-concept-representation">Human Concept Representation</h4>
<ul>
<li><p><a
href="https://www.science.org/doi/full/10.1126/science.aaf0941">Organizing
conceptual knowledge in humans with a gridlike code</a> -
<strong><em>Science</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=10995575332310321503">All
Versions</a>]. [<a
href="http://behavioralhealth2000.com/wp-content/uploads/2017/01/Organizing-conceptual-knowledge-in-humans-with-a-gridlike-code.pdf">Preprint</a>].
It has been hypothesized that the brain organizes concepts into a mental
map, allowing conceptual relationships to be navigated in a manner
similar to that of space. Grid cells use a hexagonally symmetric code to
organize spatial representations and are the likely source of a precise
hexagonal symmetry in the functional magnetic resonance imaging signal.
Humans navigating conceptual two-dimensional knowledge showed the same
hexagonal signal in a set of brain regions markedly similar to those
activated during spatial navigation. This gridlike signal is consistent
across sessions acquired within an hour and more than a week apart. This
work’s findings suggest that global relational codes may be used to
organize nonspatial conceptual representations and that these codes may
have a hexagonal gridlike pattern when conceptual knowledge is laid out
in two continuous dimensions.</p></li>
<li><p><a
href="https://www.science.org/doi/full/10.1126/science.aat6766">Navigating
cognition: Spatial codes for human thinking</a> -
<strong><em>Science</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=1407237757770081862">All
Versions</a>]. [<a
href="https://doellerlab.com/wp-content/uploads/2018/11/Bellmund_et_al_2018_Science_Navigating-cognition.pdf">Preprint</a>].
The hippocampal formation has long been suggested to underlie both
memory formation and spatial navigation. This work discusses how neural
mechanisms identified in spatial navigation research operate across
information domains to support a wide spectrum of cognitive functions.
In the proposed framework, place and grid cell population codes provide
a representational format to map variable dimensions of cognitive
spaces. This highly dynamic mapping system enables rapid reorganization
of codes through remapping between orthogonal representations across
behavioral contexts, yielding a multitude of stable cognitive spaces at
different resolutions and hierarchical levels. Action sequences result
in trajectories through cognitive space, which can be simulated via
sequential coding in the hippocampus. In this way, the spatial
representational format of the hippocampal formation has the capacity to
support flexible cognition and behavior.</p></li>
<li><p><a
href="https://www.cell.com/trends/cognitive-sciences/fulltext/S1364-6613(20)30250-3">Structuring
Knowledge with Cognitive Maps and Cognitive Graphs</a> -
<strong><em>Trends in Cognitive Sciences</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=7196012353183004425">All
Versions</a>]. [<a
href="https://www.sas.upenn.edu/psych/epsteinlab/pdfs/Peer%20Brunec%20Newcombe%20Epstein%20TiCS%202020%20Cog%20maps%20and%20cog%20graphs.pdf">Preprint</a>].
Humans and animals use mental representations of the spatial structure
of the world to navigate. The classical view is that these
representations take the form of Euclidean cognitive maps, but
alternative theories suggest that they are cognitive graphs consisting
of locations connected by paths. The authors review evidence suggesting
that both map-like and graph-like representations exist in the
mind/brain that rely on partially overlapping neural systems. Maps and
graphs can operate simultaneously or separately, and they may be applied
to both spatial and nonspatial knowledge. By providing structural
frameworks for complex information, cognitive maps and cognitive graphs
may provide fundamental organizing schemata that allow us to navigate in
physical, social, and conceptual spaces.</p></li>
<li><p><a href="https://www.nature.com/articles/nature17637">Natural
speech reveals the semantic maps that tile human cerebral cortex</a> -
<strong><em>Nature</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=14997953800741854188">All
Versions</a>]. [<a
href="https://www.polyu.edu.hk/cbs/rclcn/images/cdl_articles/H/Huth_et_al._2016.pdf">Preprint</a>].
[<a href="https://github.com/HuthLab/speechmodeltutorial">Code &amp;
Tutorial</a>]. The meaning of language is represented in regions of the
cerebral cortex collectively known as the ‘semantic system’. However,
little of the semantic system has been mapped comprehensively, and the
semantic selectivity of most regions is unknown. This work
systematically maps semantic selectivity across the cortex using
voxel-wise modelling of functional MRI (fMRI) data collected while
subjects listened to hours of narrative stories. This work shows that
the semantic system is organized into intricate patterns that seem to be
consistent across individuals. The authors then use a novel generative
model to create a detailed semantic atlas. The results suggest that most
areas within the semantic system represent information about specific
semantic domains, or groups of related concepts, and the atlas shows
which domains are represented in each area. This study demonstrates that
data-driven methods—commonplace in studies of human neuroanatomy and
functional connectivity—provide a powerful and efficient means for
mapping functional representations in the brain.</p></li>
<li><p><a
href="https://journals.sagepub.com/doi/full/10.1177/09567976211003877">Idiosyncratic
Tower of Babel: Individual differences in word-meaning representation
increase as word abstractness increases</a> - <strong><em>Psychological
Science</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=18214600097352809308">All
Versions</a>]. [<a
href="http://bilab.bnu.edu.cn/paper/2021/Wang_2021_Psychology%20Science.pdf">All
Versions</a>]. Humans primarily rely on language to communicate, on the
basis of a shared understanding of the basic building blocks of
communication: words. Do we mean the same things when we use the same
words? Although cognitive neural research on semantics has revealed the
common principles of word-meaning representation, the factors underlying
the potential individual variations in word meanings are unknown. This
work empirically characterized the intersubject consistency of 90 words
across 20 adult subjects (10 female) using both behavioral measures
(rating-based semantic-relationship patterns) and neuroimaging measures
(word-evoked brain activity patterns). Across both the behavioral and
neuroimaging experiments, this work showed that the magnitude of
individual disagreements on word meanings could be modeled on the basis
of how much language or sensory experience is associated with a word and
that this variation increases with word abstractness. Uncovering the
cognitive and neural origins of word-meaning disagreements across
individuals has implications for potential mechanisms to modulate such
disagreements.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41562-022-01316-8">Semantic
projection recovers rich human knowledge of multiple object features
from word embeddings</a> - <strong><em>Nature Human
Behavior</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=2499199921371106654">All
Versions</a>]. [<a
href="https://cap.csail.mit.edu/sites/default/files/research-pdfs/Semantic%20projection%20recovers%20rich%20human%20knowledge%20of%20multiple%20object%20features%20from%20word%20embeddings.pdf">Preprint</a>].
How is knowledge about word meaning represented in the mental lexicon?
Current computational models infer word meanings from lexical
co-occurrence patterns. They learn to represent words as vectors in a
multidimensional space, wherein words that are used in more similar
linguistic contexts—that is, are more semantically related—are located
closer together. However, whereas inter-word proximity captures only
overall relatedness, human judgements are highly context dependent. For
example, dolphins and alligators are similar in size but differ in
dangerousness. This work proposes a domain-general method to extract
context-dependent relationships from word embeddings: ‘semantic
projection’ of word-vectors onto lines that represent features such as
size (the line connecting the words ‘small’ and ‘big’) or danger (‘safe’
to ‘dangerous’), analogous to ‘mental scales’. This method recovers
human judgements across various object categories and properties. Thus,
the geometry of word embeddings explicitly represents a wealth of
context-dependent world knowledge.</p></li>
<li><p><a
href="https://www.frontiersin.org/articles/10.3389/fpsyg.2014.00385/full">Using
a high-dimensional graph of semantic space to model relationships among
words</a> - <strong><em>Frontiers in Psychology</em></strong>, 2014. [<a
href="https://scholar.google.com/scholar?cluster=472523411548302295">All
Versions</a>]. The GOLD model (Graph Of Language Distribution) is a
network model constructed based on co-occurrence in a large corpus of
natural language that may be used to explore what information may be
present in a graph-structured model of language, and what information
may be extracted through theoretically-driven algorithms as well as
standard graph analysis methods. The present study will employ GOLD to
examine two types of relationship between words: semantic similarity and
associative relatedness. Semantic similarity refers to the degree of
overlap in meaning between words, while associative relatedness refers
to the degree to which two words occur in the same schematic context. It
is expected that a graph structured model of language constructed based
on co-occurrence should easily capture associative relatedness, because
this type of relationship is thought to be present directly in lexical
co-occurrence. However, it is hypothesized that semantic similarity may
be extracted from the intersection of the set of first-order
connections, because two words that are semantically similar may occupy
similar thematic or syntactic roles across contexts and thus would
co-occur lexically with the same set of nodes. Two versions the GOLD
model that differed in terms of the co-occurence window, bigGOLD at the
paragraph level and smallGOLD at the adjacent word level, were directly
compared to the performance of a well-established distributional model,
Latent Semantic Analysis (LSA). The superior performance of the GOLD
models (big and small) suggest that a single acquisition and storage
mechanism, namely co-occurrence, can account for associative and
conceptual relationships between words and is more psychologically
plausible than models using singular value decomposition (SVD).</p></li>
<li><p><a
href="https://academic.oup.com/cercor/article/33/15/9280/7190929">Simple
shape feature computation across modalities: convergence and divergence
between the ventral and dorsal visual streams</a> - <strong><em>Cerebral
Cortex</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=5977822802446917081">All
Versions</a>]. [<a
href="http://bilab.bnu.edu.cn/paper/2023/Tian_2023_CC.pdf">Preprints</a>].
Shape processing, whether by seeing or touching, is pivotal to object
recognition and manipulation. Although the low-level signals are
initially processed by different modality-specific neural circuits,
multimodal responses to object shapes have been reported along both
ventral and dorsal visual pathways. To understand this transitional
process, the authors conducted visual and haptic shape perception fMRI
experiments to test basic shape features (i.e. curvature and
rectilinear) across the visual pathways. Using a combination of
region-of-interest-based support vector machine decoding analysis and
voxel selection method, the authors found that the top
visual-discriminative voxels in the left occipital cortex (OC) could
also classify haptic shape features, and the top haptic-discriminative
voxels in the left posterior parietal cortex (PPC) could also classify
visual shape features. Furthermore, these voxels could decode shape
features in a cross-modal manner, suggesting shared neural computation
across visual and haptic modalities. In the univariate analysis, the top
haptic-discriminative voxels in the left PPC showed haptic rectilinear
feature preference, whereas the top visual-discriminative voxels in the
left OC showed no significant shape feature preference in either of the
two modalities. Together, these results suggest that mid-level shape
features are represented in a modality-independent manner in both the
ventral and dorsal streams.</p></li>
<li><p><a href="https://www.nature.com/articles/s41597-019-0341-x">The
Database of Cross-Linguistic Colexifications, reproducible analysis of
cross-linguistic polysemies</a> - <strong><em>Scientific
Data</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=4039754406289857135">All
Versions</a>]. [<a href="https://clics.clld.org/">Project</a>]. Advances
in computer-assisted linguistic research have been greatly influential
in reshaping linguistic research. With the increasing availability of
interconnected datasets created and curated by researchers, more and
more interwoven questions can now be investigated. Such advances,
however, are bringing high requirements in terms of rigorousness for
preparing and curating datasets. This work presents CLICS, a Database of
Cross-Linguistic Colexifications (CLICS). CLICS tackles interconnected
interdisciplinary research questions about the colexifcation of words
across semantic categories in the world’s languages, and show-cases best
practices for preparing data for cross-linguistic research.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S001002772300183X">Locating
what comes to mind in empirically derived representational spaces</a> -
<strong><em>Cognition</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=57834483230365927">All
Versions</a>]. An evidence-based study concluding that people call
category members to mind according to their location in representational
space, specifically based on the predicted usefulness of considering
category members with particular features.</p></li>
<li><p><a
href="https://www.cell.com/trends/cognitive-sciences/abstract/S1364-6613(24)00171-2">Why
concepts are (probably) vectors</a> - <strong><em>Trends in Cognitive
Sciences</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=4315363807034184312">All
Versions</a>]. For decades, cognitive scientists have debated what kind
of representation might characterize human concepts. Whatever the format
of the representation, it must allow for the computation of varied
properties, including similarities, features, categories, definitions,
and relations. It must also support the development of theories, ad hoc
categories, and knowledge of procedures. Here, the authors discuss why
vector-based representations provide a compelling account that can meet
all these needs while being plausibly encoded into neural architectures.
This view has become especially promising with recent advances in both
large language models and vector symbolic architectures. These
innovations show how vectors can handle many properties traditionally
thought to be out of reach for neural models, including
compositionality, definitions, structures, and symbolic computational
processes.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="ai-concept-representation">AI Concept Representation</h4>
<ul>
<li><p><a
href="https://www.science.org/doi/full/10.1126/science.ade4401">A
principal odor map unifies diverse tasks in olfactory perception</a> -
<strong><em>Science</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=17847258457660438418">All
Versions</a>]. [<a
href="https://github.com/osmoai/publications/tree/main/lee_et_al_2023">Code</a>].
[<a
href="https://www.kaggle.com/datasets/aryanamitbarsainyan/multi-labelled-smiles-odors-dataset">Data
(Reproduced)</a>]. [<a
href="https://centaur.reading.ac.uk/113304/1/Mayhew%20et%20al%20for%20Centaur.pdf">Preprint</a>].
[<a href="https://www.thegoodscentscompany.com/">GoodScents
Database</a>]. [<a
href="http://www.leffingwell.com/bacispmp.htm">Leffingwell
Database</a>]. Mapping molecular structure to odor perception is a key
challenge in olfaction. This work used graph neural networks to generate
a principal odor map (POM) that preserves perceptual relationships and
enables odor quality prediction for previously uncharacterized odorants.
The model was as reliable as a human in describing odor quality: On a
prospective validation set of 400 out-of-sample odorants, the
model-generated odor profile more closely matched the trained panel mean
than did the median panelist. By applying simple, interpretable,
theoretically rooted transformations, the POM outperformed
chemoinformatic models on several other odor prediction tasks,
indicating that the POM successfully encoded a generalized map of
structure-odor relationships. This approach broadly enables odor
prediction and paves the way toward digitizing odors.</p></li>
<li><p><a href="https://elifesciences.org/articles/82502">Metabolic
activity organizes olfactory representations</a> -
<strong><em>eLife</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=8857896396450033667">All
Versions</a>]. [<a
href="https://github.com/osmoai/publications/tree/main/qian_et_al_2023">Code
&amp; Data</a>]. Odorous compounds with similar POM representations are
more likely to co-occur within a substance and be metabolically closely
related; metabolic reaction sequences also follow smooth paths in POM
despite large jumps in molecular structure.</p></li>
<li><p><a href="https://ieeexplore.ieee.org/abstract/document/9136877">A
Review of Tactile Information: Perception and Action Through Touch</a> -
<strong><em>IEEE Transactions on Robotics</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=15493221881484741343">All
Versions</a>]. [<a
href="https://www.researchgate.net/profile/Qiang-Li-110/publication/342797645_A_Review_of_Tactile_Information_Perception_and_Action_Through_Touch/links/602f95bc92851c4ed5806e9f/A-Review-of-Tactile-Information-Perception-and-Action-Through-Touch.pdf">Preprint</a>].
Tactile sensing is a key sensor modality for robots interacting with
their surroundings. These sensors provide a rich and diverse set of data
signals that contain detailed information collected from contacts
between the robot and its environment. The data are however not limited
to individual contacts and can be used to extract a wide range of
information about the objects in the environment as well as the actions
of the robot during the interactions. This article provides an overview
of tactile information and its applications in robotics. The authors
present a hierarchy consisting of raw, contact, object, and action
levels to structure the tactile information, with higher-level
information often building upon lower-level information. The authors
discuss different types of information that can be extracted at each
level of the hierarchy. The article also includes an overview of
different types of robot applications and the types of tactile
information that they employ. Finally the article ends with a discussion
for future tactile applications which are still beyond the current
capabilities of robots.</p></li>
<li><p><a
href="https://openaccess.thecvf.com/content/CVPR2023/html/Girdhar_ImageBind_One_Embedding_Space_To_Bind_Them_All_CVPR_2023_paper.html">ImageBind:
One Embedding Space To Bind Them All</a> -
<strong><em>CVPR’23</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=1657173986906232916">All
Versions</a>]. [<a
href="https://github.com/facebookresearch/ImageBind">Project</a>]. This
work presents ImageBind, an approach to learn a joint embedding across
six different modalities - images, text, audio, depth, thermal, and IMU
data. The authors show that all combinations of paired data are not
necessary to train such a joint embedding, and only image-paired data is
sufficient to bind the modalities together. ImageBind can leverage
recent large scale vision-language models, and extends their zero-shot
capabilities to new modalities just by using their natural pairing with
images. It enables novel emergent applications ‘out-of-the-box’
including cross-modal retrieval, composing modalities with arithmetic,
cross-modal detection and generation. The emergent capabilities improve
with the strength of the image encoder and this work sets a new
state-of-the-art on emergent zero-shot recognition tasks across
modalities, outperforming specialist supervised models. Finally, the
authors show strong few-shot recognition results outperforming prior
work, and that ImageBind serves as a new way to evaluate vision models
for visual and non-visual tasks.</p></li>
<li><p><a href="https://escholarship.org/uc/item/44s454ng">Semantic
features of object concepts generated with GPT-3</a> -
<strong><em>CogSci’22</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=16958563995984242923">All
Versions</a>]. Semantic features have been playing a central role in
investigating the nature of our conceptual representations. Yet the
enormous time and effort required to empirically sample and norm
features from human raters has restricted their use to a limited set of
manually curated concepts. Given recent promising developments with
transformer-based language models, here the authors asked whether it was
possible to use such models to automatically generate meaningful lists
of properties for arbitrary object concepts and whether these models
would produce features similar to those found in humans. To this end,
the authors probed a GPT-3 model to generate semantic features for 1,854
objects and compared automatically-generated features to existing human
feature norms. GPT-3 generated many more features than humans, yet
showed a similar distribution in the types of generated features.
Generated feature norms rivaled human norms in predicting similarity,
relatedness, and category membership, while variance partitioning
demonstrated that these predictions were driven by similar variance in
humans and GPT-3. Together, these results highlight the potential of
large language models to capture important facets of human knowledge and
yield a new approach for automatically generating interpretable feature
sets, thus drastically expanding the potential use of semantic features
in psychological and linguistic studies.</p></li>
<li><p><a href="https://ieeexplore.ieee.org/document/8953737">Connecting
Touch and Vision via Cross-Modal Prediction</a> -
<strong><em>CVPR’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=17326564895972374001">All
Versions</a>]. [<a
href="https://github.com/YunzhuLi/VisGel">Project</a>]. Humans perceive
the world using multi-modal sensory inputs such as vision, audition, and
touch. This work investigates the cross-modal connection between vision
and touch. The main challenge in this cross-domain modeling task lies in
the significant scale discrepancy between the two: while our eyes
perceive an entire visual scene at once, humans can only feel a small
region of an object at any given moment. To connect vision and touch,
this work introduces new tasks of synthesizing plausible tactile signals
from visual inputs as well as imagining how we interact with objects
given tactile data as input. To accomplish the goals, the authors first
equip robots with both visual and tactile sensors and collect a
large-scale dataset of corresponding vision and tactile image sequences.
To close the scale gap, the authors present a new conditional
adversarial model that incorporates the scale and location information
of the touch. Human perceptual studies demonstrate that the model can
produce realistic visual images from tactile data and vice
versa.</p></li>
<li><p><a href="https://aclanthology.org/2022.tacl-1.69/">Unit Testing
for Concepts in Neural Networks</a> - <strong><em>Transactions of the
Association for Computational Linguistics</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=3036662275506971282&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Testing the concept representation by neural networks
through Fodor’s theory of concepts.</p></li>
<li><p><a href="https://aclanthology.org/2024.acl-long.820/">Do Llamas
Work in English? On the Latent Language of Multilingual Transformers</a>
- <strong><em>ACL’24</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=5847238732288003106">All
Versions</a>]. A preliminary work empirically showing that the
intermediate embeddings of multilingual Transformers (1) start far away
from output token embeddings; (2) already allow for decoding a
semantically correct next token in the middle layers, but give higher
probability to its version in English than in the input language; (3)
finally move into an input-language-specific region of the embedding
space. Also, the embedding of abstract concept space lies closer to
English than to other languages.</p></li>
<li><p><a
href="https://www.cell.com/trends/cognitive-sciences/abstract/S1364-6613(24)00035-4">From
task structures to world models: what do LLMs know?</a> -
<strong><em>Trends in Cognitive Sciences</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=14836877410607949822">All
Versions</a>]. [<a
href="http://cncl.yale.edu/sites/default/files/pub-downloads/yildirim-paul-llms-knowledge.pdf">Preprint</a>].
In what sense does a large language model (LLM) have knowledge? The
authors answer by granting LLMs ‘instrumental knowledge’: knowledge
gained by using next-word generation as an instrument. The authors then
ask how instrumental knowledge is related to the ordinary, ‘worldly
knowledge’ exhibited by humans, and explore this question in terms of
the degree to which instrumental knowledge can be said to incorporate
the structured world models of cognitive science. The authors discuss
ways LLMs could recover degrees of worldly knowledge and suggest that
such recovery will be governed by an implicit, resource-rational
tradeoff between world models and tasks. The authors’ answer to this
question extends beyond the capabilities of a particular AI system and
challenges assumptions about the nature of knowledge and
intelligence.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="complexity-information-theory">Complexity &amp; Information
Theory</h3>
<h4 id="theory">Theory</h4>
<ul>
<li><p><a
href="http://www.cs.yale.edu/homes/yry/readings/general/shannon1948.pdf">A
Mathematical Theory of Communication</a> - <strong><em>The Bell System
Technical Journal</em></strong>, 1948. [<a
href="https://scholar.google.com/scholar?cluster=8313213127749369813">All
Versions</a>]. Shannon’s original paper on Information Theory.</p></li>
<li><p><a
href="https://link.springer.com/content/pdf/10.1007/978-3-030-11298-1.pdf">An
introduction to Kolmogorov complexity and its applications</a> -
<strong><em>Springer</em></strong>, 2008. [<a
href="https://scholar.google.com/scholar?cluster=8746708322477453221">All
Versions</a>]. The introductory book for Algorithmic Information Theory,
especially the Kolmogorov complexity theory.</p></li>
<li><p><a
href="https://psycnet.apa.org/record/1973-01647-001">Complexity and the
representation of patterned sequences of symbols</a> -
<strong><em>Psychological Review</em></strong>, 1972. [<a
href="https://scholar.google.com/scholar?cluster=3426861135318645138">All
Versions</a>]. Herbert Simon’s review on subjective complexity.</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/abstract/document/1057698">Visual
Pattern Discrimination</a> - <strong><em>IRE Transactions on Information
Theory</em></strong>, 1962. [<a
href="https://scholar.google.com/scholar?cluster=10729525966103382864">All
Versions</a>].</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&amp;arnumber=5390997">Algorithmic
Information Theory</a> - <strong><em>IBM Journal of Research and
Development</em></strong>, 1977. [<a
href="https://scholar.google.com/scholar?cluster=14735710867906424793">All
Versions</a>]. Chaitin’s original paper on Algorithmic Information
Theory.</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2003/hash/b06b5541a62ed438f956b662b4e1ec28-Abstract.html">From
Algorithmic to Subjective Randomness</a> -
<strong><em>NeurIPS’03</em></strong>, 2003. [<a
href="https://scholar.google.com/scholar?cluster=14721764738308036578">All
Versions</a>].</p></li>
<li><p><a href="https://proceedings.mlr.press/v202/shi23i.html">On the
Complexity of Bayesian Generalization</a> -
<strong><em>ICML’23</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=5817813824878811147">All
Versions</a>]. [<a
href="https://github.com/SHI-Yu-Zhe/bayesian-generalization-complexity">Project</a>].
[<a
href="https://drive.google.com/file/d/1eCuFqBYN8kuiAmoVtXWedXW0r0TdY55W/view">Models</a>].
This work examines concept generalization at a large scale in the
natural visual spectrum. Established computational modes (i.e.,
rule-based or similarity-based) are primarily studied isolated, focusing
on confined and abstract problem spaces. This work studies these two
modes when the problem space scales up and when the complexity of
concepts becomes diverse. At the representational level, the authors
investigate how the complexity varies when a visual concept is mapped to
the representation space. Prior literature has shown that two types of
complexities build an inverted-U relation. Leveraging Representativeness
of Attribute (RoA), the authors computationally confirm: Models use
attributes with high RoA to describe visual concepts, and the
description length falls in an inverted-U relation with the increment in
visual complexity. At the computational level, the authors examine how
the complexity of representation affects the shift between the rule- and
similarity-based generalization. The authors hypothesize that
category-conditioned visual modeling estimates the co-occurrence
frequency between visual and categorical attributes, thus potentially
serving as the prior for the natural visual world. Experimental results
show that representations with relatively high subjective complexity
outperform those with relatively low subjective complexity in rule-based
generalization, while the trend is the opposite in similarity-based
generalization.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="dimensionality-reduction">Dimensionality Reduction</h4>
<ul>
<li><p><a
href="https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.1084.4695&amp;rep=rep1&amp;type=pdf">A
global geometric framework for nonlinear dimensionality reduction</a> -
<strong><em>Science</em></strong>, 2000. [<a
href="https://scholar.google.com/scholar?cluster=14602426245887619907">All
Versions</a>]. The original paper on spectrum clustering.</p></li>
<li><p><a
href="https://asset-pdf.scinapse.io/prod/2100495367/2100495367.pdf">Reducing
the dimensionality of data with neural networks</a> -
<strong><em>Science</em></strong>, 2006. [<a
href="https://scholar.google.com/scholar?cluster=15344645275208957628">All
Versions</a>]. The original paper on Variational Autoencoder.</p></li>
<li><p><a href="https://arxiv.org/pdf/1206.5538.pdf">Representation
Learning: A Review and New Perspectives</a> - <strong><em>IEEE
Transactions on Pattern Analysis and Machine Intelligence</em></strong>,
2013. [<a
href="https://scholar.google.com/scholar?cluster=559463397382443088">All
Versions</a>]. Yoshua Bengio’s review on representation
learning.</p></li>
<li><p><a
href="http://www.stat.ucla.edu/~jxie/personalpage_file/publications/representation_learning_Review.pdf">Representation
Learning: A Statistical Perspective</a> - <strong><em>Annual Review of
Statistics and Its Application</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=14358027809538175293">All
Versions</a>]. Song-Chun Zhu and Ying Nian Wu’s review on representation
learning, in an account of statistics.</p></li>
<li><p><a
href="http://robotics.caltech.edu/wiki/images/8/8f/DeepLearningBottleneck.pdf">Deep
Learning and the Information Bottleneck Principle</a> - <strong><em>IEEE
Information Theory Workshop’15</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=13152354842433826281">All
Versions</a>]. The first paper identifying the problem of information
bottleneck in representation learning.</p></li>
<li><p><a
href="https://artemyk.github.io/assets/pdf/papers/Saxe%20et%20al_2019_On%20the%20information%20bottleneck%20theory%20of%20deep%20learning.pdf">On
the information bottleneck theory of deep learning</a> -
<strong><em>Journal of Statistical Mechanics: Theory and
Experiment</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=12271240925674881982">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="visual-complexity">Visual Complexity</h4>
<ul>
<li><p><a
href="https://www.researchgate.net/profile/Don-Donderi-2/publication/7337589_Visual_Complexity_A_Review/links/5f0875ed45851550509a3a7a/Visual-Complexity-A-Review.pdf">Visual
complexity: a review</a> - <strong><em>Psychological
Bulletin</em></strong>, 2006. [<a
href="https://scholar.google.com/scholar?cluster=10747901143387624939">All
Versions</a>]. [<a
href="https://psycnet.apa.org/record/2006-00818-005">APA</a>]. A
psychological account on visual complexity.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0141938205000120">Compressed
File Length Predicts Search Time and Errors on Visual Displays</a> -
<strong><em>Displays</em></strong>, 2005. [<a
href="https://scholar.google.com/scholar?cluster=15600966633648834042">All
Versions</a>]. Compressed file size, an objective, easily obtained
measure of display complexity, predicts both subjective complexity
judgments and objective search performance. It is analogous to
algorithmic complexity, a theoretical but impractical measure of bit
string complexity. The data suggest that it may be possible to use the
compressed file size measure to predict display performance in applied
tasks.</p></li>
<li><p><a
href="https://stefan.winklerbros.net/Publications/qomex2013si.pdf">Image
complexity and spatial information</a> - <strong><em>International
Workshop on Quality of Multimedia Experience</em></strong>, 2013. [<a
href="https://scholar.google.com/scholar?cluster=16011036229039693102">All
Versions</a>].</p></li>
<li><p><a
href="https://perception.jhu.edu/files/PDFs/21_Complexity_Speaking/SunFirestone_SpeakingSeeing_2021_JEPG.pdf">Seeing
and speaking: How verbal “description length” encodes visual
complexity</a> - <strong><em>Journal of Experimental
Psychology</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=246820603191585233">All
Versions</a>]. [<a
href="https://psycnet.apa.org/record/2021-83037-001">APA</a>]. Empirical
evidencs showing the relation between visual complexity and description
length.</p></li>
<li><p><a
href="https://pure.mpg.de/rest/items/item_3380375/component/file_3383568/content">How
variability shapes learning and generalization</a> - <strong><em>Trends
in Cognitive Sciences</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=10940775338620708972">All
Versions</a>]. A comprehensive review on the trade-off between
variability and generalization ability.</p></li>
<li><p><a href="https://arxiv.org/pdf/2205.05666.pdf">Identifying
concept libraries from language about object structure</a> -
<strong><em>CogSci’22</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=4019205027627496528">All
Versions</a>].</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0010027722003158">Show
or tell? Exploring when (and why) teaching with language outperforms
demonstration</a> - <strong><em>Cognition</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=11837154580063293174">All
Versions</a>]. The findings of this paper suggest that language
communicates complex concepts by directly transmitting abstract rules.
In contrast, demonstrations transmit examples, requiring the learner to
infer the rules.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="communications">Communications</h3>
<h4 id="non-verbal-communication">Non-Verbal Communication</h4>
<ul>
<li><p><a
href="https://onlinelibrary.wiley.com/doi/epdf/10.1111/j.1551-6709.2009.01090.x">The
Interactive Evolution of Human Communication Systems</a> -
<strong><em>Cognitive Science</em></strong>, 2010. [<a
href="https://scholar.google.com/scholar?cluster=6689941517686043970">All
Versions</a>]. Nicolas Fay’s original paper on iconicity.</p></li>
<li><p><a href="https://benjamins.com/catalog/pc.22.2.05fay">Iconicity:
From sign to system in human communication and language</a> -
<strong><em>Pragmatics &amp; Cognition</em></strong>, 2014. [<a
href="https://scholar.google.com/scholar?cluster=8525760321117094567">All
Versions</a>]. This paper explores the role of iconicity in spoken
language and other human communication systems.</p></li>
<li><p><a
href="https://journals.sagepub.com/doi/abs/10.1177/108835769400900301">The
Picture Exchange Communication System</a> - <strong><em>Behavior
Modification</em></strong>, 1994. [<a
href="https://scholar.google.com/scholar?cluster=18113491434570143349&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://onlinelibrary.wiley.com/doi/full/10.1080/15326900701221363">Graphical
Language Games: Interactional Constraints on Representational Form</a> -
<strong><em>Cognitive Science</em></strong>, 2007. [<a
href="https://scholar.google.com/scholar?cluster=280214578402050136&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The first paper introducing the graphical language
game.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0378216614001830">A
multimodal discourse theory of visual narrative</a> -
<strong><em>Journal of Pragmatics</em></strong>, 2014. [<a
href="https://scholar.google.com/scholar?cluster=912273653379961242&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://ayankumarbhunia.github.io/pixelor/image/pixelor.pdf">Pixelor:
A Competitive Sketching AI Agent. So you think you can beat me?</a> -
<strong><em>ACM SIGGRAPH’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=6676723059377806081&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a href="http://sketchx.ai/pixelor">Project</a>].
Rationality in feature sketching.</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007/s42113-019-00058-7">Pragmatic
Inference and Visual Abstraction Enable Contextual Flexibility During
Visual Communication</a> - <strong><em>Computational Brain &amp;
Behavior</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=17971107104483505071&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A computational account on the rational behavior in
graphical language games.</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper_files/paper/2022/hash/550ff553efc2c58410f277c667d12786-Abstract-Conference.html">Emergent
Graphical Conventions in a Visual Communication Game</a> -
<strong><em>NeurIPS</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=17122070906194572150">All
Versions</a>]. A computational account on the emergence of iconic
language.</p></li>
<li><p><a href="https://dl.acm.org/doi/abs/10.1145/3610591.3616427">AI
Nüshu: An Exploration of Language Emergence in Sisterhood Through the
Lens of Computational Linguistics</a> - <strong><em>ACM SIGGRAPH
Asia’23</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=6849286654402017109&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. By continually observing their environment and
communicating, AI agents trained in the Chinese dictionary and the Nüshu
corpus collaborate towards creating a standard writing system to encode
Chinese.</p></li>
<li><p><a
href="https://www.pnas.org/content/118/12/e2016569118">Communicating
artificial neural networks develop efficient color-naming systems</a> -
<strong><em>Proceedings of the National Academy of
Sciences</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=1640459156303560508&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Simulating the emergence of code as the communication
bottleneck in color learning task.</p></li>
<li><p><a
href="https://escholarship.org/content/qt9p70d5s9/qt9p70d5s9.pdf">Bridging
cultural and cognitive perspectives on similarity reasoning</a> -
<strong><em>CogSci’22</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?hl=en&amp;as_sdt=0%2C5&amp;q=Bridging+cultural+and+cognitive+perspectives+on+similarity+reasoning&amp;btnG=">All
Versions</a>].</p></li>
<li><p><a
href="https://www.eva.mpg.de/documents/Elsevier/Liszkowski_Twelve_Cognition_2008_1554509.pdf">Twelve-month-olds
communicate helpfully and appropriately for knowledgeable and ignorant
partners</a> - <strong><em>Cognition</em></strong>, 2008. [<a
href="https://scholar.google.com/scholar?cluster=8202048572661677635&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper on child pointing.</p></li>
<li><p><a
href="https://pure.mpg.de/rest/items/item_64467_4/component/file_64468/content">12-
and 18-Month-Olds Point to Provide Information for Others</a> -
<strong><em>Journal of Cognition and Development</em></strong>, 2009.
[<a
href="https://scholar.google.com/scholar?cluster=7322772656439413984&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007/s10115-006-0062-2">Toward
understanding the importance of gesture in distributed scientific
collaboration</a> - <strong><em>Knowledge and Information
Systems</em></strong>, 2006. [<a
href="https://scholar.google.com/scholar?cluster=3145646721897130511">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="pragmatics">Pragmatics</h4>
<ul>
<li><p><a
href="https://plato.stanford.edu/entries/pragmatics/">Pragmatics</a> -
<strong><em>Plato Stanford</em></strong>. A computational philosophy
account of Pragmatics, whilch studies utterances in specific
contexts.</p></li>
<li><p><a
href="https://www.science.org/doi/abs/10.1126/science.1218633">Predicting
Pragmatic Reasoning in Language Games</a> -
<strong><em>Science</em></strong>, 2012. [<a
href="https://scholar.google.com/scholar?cluster=15533081031935746054">All
Versions</a>]. [<a
href="https://langcog.stanford.edu/papers_new/frank-2012-science.pdf">Preprint</a>].
One of the most astonishing features of human language is its capacity
to convey information efficiently in context. Many theories provide
informal accounts of communicative inference, yet there have been few
successes in making precise, quantitative predictions about pragmatic
reasoning. This work examined judgments about simple referential
communication games, modeling behavior in these games by assuming that
speakers attempt to be informative and that listeners use Bayesian
inference to recover speakers’ intended referents. The model provides a
close, parameter-free fit to human judgments, suggesting that the use of
information-theoretic tools to predict pragmatic reasoning may lead to
more effective formal models of communication.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S136466131630122X">Pragmatic
Language Interpretation as Probabilistic Inference</a> -
<strong><em>Trends in Cognitive Sciences</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=11393505968563356130">All
Versions</a>]. Understanding language requires more than the use of
fixed conventions and more than decoding combinatorial structure.
Instead, comprehenders make exquisitely sensitive inferences about what
utterances mean given their knowledge of the speaker, language, and
context. Building on developments in game theory and probabilistic
modeling, the authors describe the rational speech act (RSA) framework
for pragmatic reasoning. RSA models provide a principled way to
formalize inferences about meaning in context; they have been used to
make successful quantitative predictions about human behavior in a
variety of different tasks and situations, and they explain why complex
phenomena, such as hyperbole and vagueness, occur. More generally, they
provide a computational framework for integrating linguistic structure,
world knowledge, and context in pragmatic language
understanding.</p></li>
<li><p><a
href="http://cocolab.stanford.edu/papers/BergenLevyGoodman-LexUnc.pdf">Pragmatic
Reasoning through Semantic Inference</a> - <strong><em>Semantics &amp;
Pragmatics</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=1433855075217315997">All
Versions</a>].</p></li>
<li><p><a
href="https://semantics.uchicago.edu/kennedy/docs/processing.pdf">Processing
gradable adjectives in context: A visual world study</a> -
<strong><em>Semantics and Linguistic Theory</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=13426776838629402579">All
Versions</a>]. Adjective understanding as a rational inference in the
context.</p></li>
<li><p><a
href="https://transacl.org/index.php/tacl/article/view/1142">Colors in
Context: A Pragmatic Neural Model for Grounded Language
Understanding</a> - <strong><em>Transactions of the Association for
Computational Linguistics</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=11119271811833503059">All
Versions</a>].</p></li>
<li><p><a
href="https://compdevlab.yale.edu/docs/2019/2019_ChildDev_Pragmatics.pdf">Social
Pragmatics: Preschoolers Rely on Commonsense Psychology to Resolve
Referential Underspecification</a> - <strong><em>Child
Development</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=16352913537004112920">All
Versions</a>]. A piece of evidence for children’s capability on social
pragmatics.</p></li>
<li><p><a
href="http://cocolab.stanford.edu/papers/CohnGordonEtAl2018_NAACL.pdf">Pragmatically
Informative Image Captioning with Character-Level Inference</a> -
<strong><em>NAACL’18</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=1670953084401884599">All
Versions</a>].</p></li>
<li><p><a
href="https://aclanthology.org/2020.findings-emnlp.173/">Pragmatic
Issue-Sensitive Image Captioning</a> - <strong><em>EMNLP
Findings’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=10608257248144445301">All
Versions</a>]. Application of Rational Speech Act to Image
Captioning.</p></li>
<li><p><a
href="https://cogsci.mindmodeling.org/2019/papers/0091/0091.pdf">Disentangling
contributions of visual information and interaction history in the
formation of graphical conventions</a> -
<strong><em>CogSci’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=15046353579508199394&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://www.nature.com/articles/s41562-021-01145-1">How
young children integrate information sources to infer the meaning of
words</a> - <strong><em>Nature Human Behavior</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=10144794357802769844">All
Versions</a>]. Before formal education begins, children typically
acquire a vocabulary of thousands of words. This learning process
requires the use of many different information sources in their social
environment, including their current state of knowledge and the context
in which they hear words used. This paper specifies a developmental
model according to which children consider information sources in an
age-specific way and integrate them via Bayesian inference. This work
presents a developmental theory of information integration during
language learning and illustrates how formal models can be used to make
a quantitative test of the predictive and explanatory power of competing
theories.</p></li>
<li><p><a
href="https://semprag.org/index.php/sp/article/view/sp.5.6/pdf">Information
Structure in Discourse: Towards an Integrated Formal Theory of
Pragmatics</a> - <strong><em>Semantics and Pragmatics</em></strong>,
1998. [<a
href="https://scholar.google.com/scholar?cluster=9127222314768938599&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007/s11098-020-01490-3">When
Lingens meets Frege: communication without common ground</a> -
<strong><em>Philosophical Studies</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=10912415595149303257&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/2307.07871">The SocialAI School:
Insights from Developmental Psychology Towards Artificial Socio-Cultural
Agents</a> - <strong><em>ICML’23 Workshop on
Theory-of-Mind</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=11933410239580707313&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://sites.google.com/view/socialai-school">Project</a>].</p></li>
<li><p><a
href="https://www.nature.com/articles/s41599-020-0404-9">Language as
shaped by the environment: linguistic construal in a collaborative
spatial task</a> - <strong><em>Humanities and Social Sciences
Communications</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=7842508027049437987">All
Versions</a>]. [<a href="https://osf.io/sxtaq">Code &amp; Data</a>]. [<a
href="https://dialoguetoolkit.github.io/chattool/">Dialogue Experimental
Toolkit(DiET)</a>]. The present study sets out to experimentally
investigate how environmental factors come to shape the emergence of
linguistic conventions. To this end, the authors adapt the classical
Maze Game task to test the hypothesis that participants routinise
different linguistic strategies to communicate positions in the maze
contingent on particular environmental affordances (i.e. structure of
the mazes). The results confirm that subtle environmental motivations
drive the emergence of different communicative conventions in an
otherwise identical task, suggesting that linguistic adaptations are
highly sensitive to factors of the shared task environment.</p></li>
<li><p><a href="https://arxiv.org/abs/2008.12142">Exploring Urban Form
Through Openstreetmap Data: A Visual Introduction</a> -
<strong><em>Urban Experience and Design: Contemporary Perspectives on
Improving the Public Realm</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=7094530618542001733&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a href="https://github.com/gboeing/osmnx">OSMnx
Tool</a>]. [<a href="https://www.openstreetmap.org/">OpenStreetMap
Website</a>].</p></li>
<li><p><a
href="https://www.speech.kth.se/~edlund/bielefeld/references/garrod-and-anderson-1987.pdf">Saying
what you mean in dialogue: A study in conceptual and semantic
co-ordination</a> - <strong><em>Cognition</em></strong>, 1987. [<a
href="https://scholar.google.com/scholar?cluster=15377075954534820544&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://www.sfs.uni-tuebingen.de/~gjaeger/lehre/ws0708/spieltheorie/garrod.pdf">Conversation,
co-ordination and convention: an empirical investigation of how groups
establish linguistic conventions</a> -
<strong><em>Cognition</em></strong>, 1994. [<a
href="https://scholar.google.com/scholar?cluster=3784850469297049700&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="language-compositionality">Language Compositionality</h4>
<ul>
<li><p><a
href="https://plato.stanford.edu/entries/compositionality/">Compositionality</a>
- <strong><em>Plato Stanford</em></strong>. A computational philosophy
account on compositionality, one of the distinctive feature of
language.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41586-024-07522-w">Language is
primarily a tool for communication rather than thought</a> -
<strong><em>Nature</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=13724799649075764503">All
Versions</a>]. This perspective brings recent evidence from neuroscience
and allied disciplines to argue that in modern humans, language is a
tool for communication, contrary to a prominent view that we use
language for thinking. The authors begins by introducing the brain
network that supports linguistic ability in humans. They then review
evidence for a double dissociation between language and thought, and
discuss several properties of language that suggest that it is optimized
for communication. This perspective concludes that although the
emergence of language has unquestionably transformed human culture,
language does not appear to be a prerequisite for complex thought,
including symbolic thought. Instead, language is a powerful tool for the
transmission of cultural knowledge; it plausibly co-evolved with humans’
thinking and reasoning capacities, and only reflects, rather than gives
rise to, the signature sophistication of human cognition.</p></li>
<li><p><a
href="https://link.springer.com/content/pdf/10.1007/BF00763644.pdf">The
Principle of Semantic Compositionality</a> -
<strong><em>Topoi</em></strong>, 1994. [<a
href="https://scholar.google.com/scholar?cluster=10899040818001759322&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper on the principle of semantic
compositionality.</p></li>
<li><p><a
href="http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.60.3235">On
The Emergence Of Compositionality</a> - <strong><em>Proceedings of the
Evolution of Language Conference’06</em></strong>, 2006. [<a
href="https://scholar.google.com/scholar?cluster=16315741180717951222&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper on the emergence of
compositionality.</p></li>
<li><p><a href="https://arxiv.org/pdf/1612.07182.pdf">Multi-Agent
Cooperation and the Emergence of (Natural) Language</a> -
<strong><em>ICLR’17</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=1931070702879918446&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper on the emergence of language in
multi-agent reinforcement learning.</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2017/hash/70222949cc0db89ab32c9969754d4758-Abstract.html">Emergence
of Language with Multi-agent Games: Learning to Communicate with
Sequences of Symbols</a> - <strong><em>NeurIPS’18</em></strong>, 2018.
[<a
href="https://scholar.google.com/scholar?cluster=17308624474306270808&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/1804.03980">Emergent communication
through negotiation</a> - <strong><em>ICLR’18</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=8825869866742501521&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://psycnet.apa.org/record/2019-07481-001">The
language of generalization</a> - <strong><em>Psychological
Review</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=7723877614160376324&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/2004.09124">Compositionality and
Generalization in Emergent Languages</a> -
<strong><em>ACL’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=5792073344743965767&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://escholarship.org/uc/item/5kv636c5">Word
formation supports efficient communication: The case of compounds</a> -
<strong><em>CogSci’22</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=17465553221758916299&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/2311.17227">War and Peace
(WarAgent): Large Language Model-based Multi-Agent Simulation of World
Wars</a> - 2023. [<a
href="https://scholar.google.com/scholar?cluster=3598519753107761968&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="coordination">Coordination</h4>
<ul>
<li><p><a
href="https://www.science.org/doi/full/10.1126/scirobotics.abm4183">In
situ bidirectional human-robot value alignment</a> - <strong><em>Science
Robotics</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=18342339995965564405">All
Versions</a>]. [<a
href="https://par.nsf.gov/servlets/purl/10351399">Preprint</a>]. This
paper proposes an explainable artificial intelligence (XAI) system in
which a group of robots predicts users’ values by taking in situ
feedback into consideration while communicating their decision processes
to users through explanations. To learn from human feedback, the XAI
system integrates a cooperative communication model for inferring human
values associated with multiple desirable goals. To be interpretable to
humans, it simulates human mental dynamics and predicts optimal
explanations using graphical models.</p></li>
<li><p><a href="https://arxiv.org/pdf/2304.14656.pdf">From Explicit
Communication to Tacit Cooperation: A Novel Paradigm for Cooperative
MARL</a> - <strong><em>AAMAS’24</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=12114270828108588849">All
Versions</a>]. Drawing inspiration from human team cooperative learning,
this paper proposes a novel paradigm that facilitates a gradual shift
from explicit communication to tacit cooperation.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="domain-specific-language">Domain Specific Language</h3>
<h4 id="design-theory">Design Theory</h4>
<ul>
<li><p><a
href="https://en.wikipedia.org/wiki/Domain-specific_language">Domain-Specific
Language</a> - <strong><em>Wikipedia</em></strong>. Wikipedia
encyclopedia entry on Domain Specific Languages.</p></li>
<li><p><a href="https://en.wikipedia.org/wiki/Domain_engineering">Domain
Engineering</a> - <strong><em>Wikipedia</em></strong>. Wikipedia
encyclopedia entry on Domain Engineering.</p></li>
<li><p><a href="https://martinfowler.com/books/dsl.html">Domain-Specific
Languages</a> - <strong><em>Pearson Education</em></strong>, 2010. [<a
href="https://scholar.google.com/scholar?cluster=3653365103385845410">All
Versions</a>]. [<a
href="https://martinfowler.com/dsl.html">Domain-Specific Languages
Guide</a>]. When carefully selected and used, Domain-Specific Languages
(DSLs) may simplify complex code, promote effective communication with
customers, improve productivity, and unclog development bottlenecks. In
Domain-Specific Languages, noted software development expert Martin
Fowler first provides the information software professionals need to
decide if and when to utilize DSLs. Then, where DSLs prove suitable,
Fowler presents effective techniques for building them, and guides
software engineers in choosing the right approaches for their
applications.</p></li>
<li><p><a
href="https://en.wikipedia.org/wiki/Comparison_of_multi-paradigm_programming_languages">Comparison
of multi-paradigm programming languages</a> -
<strong><em>Wikipedia</em></strong>. Programming languages may support
multiple programming paradigms. This Wikipedia encyclopedia entry lists
a concise reference for the programming paradigms.</p></li>
<li><p><a
href="https://dl.acm.org/doi/pdf/10.1145/947955.1083808">Epigrams on
programming</a> - <strong><em>ACM SIGPLAN Notices</em></strong>, 1982.
[<a
href="https://scholar.google.com/scholar?cluster=6439127299132936476">All
Versions</a>].</p></li>
<li><p><a href="https://tomassetti.me/domain-specific-languages/">The
complete guide to (external) Domain Specific Languages</a>. An
introduction to Domain Specific Languages (DSL) based on 19 DSL
cases.</p></li>
<li><p><a href="https://dl.acm.org/doi/abs/10.1145/1118890.1118892">When
and How to Develop Domain-Specific Languages</a> - <strong><em>ACM
Computing Surveys</em></strong>, 2005. [<a
href="https://scholar.google.com/scholar?cluster=8598236436890577027">All
Versions</a>]. [<a
href="https://people.cs.ksu.edu/~schmidt/505f14/Lectures/WhenDSL.pdf">Preprint</a>].
Domain-specific languages (DSLs) are languages tailored to a specific
application domain. They offer substantial gains in expressiveness and
ease of use compared with general-purpose programming languages in their
domain of application. DSL development is hard, requiring both domain
knowledge and language development expertise. Few people have both. Not
surprisingly, the decision to develop a DSL is often postponed
indefinitely, if considered at all, and most DSLs never get beyond the
application library stage. Although many articles have been written on
the development of particular DSLs, there is very limited literature on
DSL development methodologies and many questions remain regarding when
and how to develop a DSL. To aid the DSL developer, this survey paper
identifies patterns in the decision, analysis, design, and
implementation phases of DSL development. These patterns improve and
extend earlier work on DSL design patterns.</p></li>
<li><p><a href="https://arxiv.org/abs/1409.2378">Design Guidelines for
Domain Specific Languages</a> - <strong><em>OOPSLA Workshop on
Domain-Specific Modeling (DSM’ 09)</em></strong>, 2009. [<a
href="https://scholar.google.com/scholar?cluster=1962567819031018744">All
Versions</a>]. Designing a new domain specific language is as any other
complex task sometimes error-prone and usually time consuming,
especially if the language shall be of high-quality and comfortably
usable. Existing tool support focuses on the simplification of technical
aspects but lacks support for an enforcement of principles for a good
language design. In this paper we investigate guidelines that are useful
for designing domain specific languages, largely based on our experience
in developing languages as well as relying on existing guidelines on
general purpose (GPLs) and modeling languages. This work defined
Guidelines to support a DSL developer to achieve better quality of the
language design and a better acceptance among its users.</p></li>
<li><p><a
href="https://dl.acm.org/doi/abs/10.1145/352029.352035">Domain-specific
languages: an annotated bibliography</a> - <strong><em>ACM SIGPLAN
Notices</em></strong>, 2000. [<a
href="https://scholar.google.com/scholar?cluster=8845429548327315750">All
Versions</a>]. A survey on the topic of domain-specific languages as
used for the construction and maintenance of software systems. The
survey lists a selection of 75 key publications in the area, and
provides a summary for each of the papers. Moreover, the survey
discusses terminology, risks and benefits, example domain-specific
languages, design methodologies, and implementation techniques.</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/abstract/document/6511840">Usability
Evaluation of Domain-Specific Languages</a> -
<strong><em>ICQICT’12</em></strong>, 2012. [<a
href="https://scholar.google.com/scholar?cluster=3047215455890195199">All
Versions</a>]. [<a
href="http://www-ctp.di.fct.unl.pt/QUASAR/Resources/Papers/2012/Barisic2012SEDES.pdf">Preprint</a>].
The purpose of this proposal is to contribute to the systematic activity
of Software Language Engineering by focusing on the issue of the
Usability evaluation of DSLs. Usability evaluation is often skipped,
relaxed, or at least omitted from papers reporting development of DSLs.
The authors argue that a systematic approach based on User Interface
experimental validation techniques should be used to assess the impact
of new DSLs. For that purpose, the authors propose to merge common
Usability evaluation processes with the DSL development
process.</p></li>
<li><p><a
href="https://link.springer.com/chapter/10.1007/978-3-642-36654-3_6">Domain-Specific
Modeling Languages: Requirements Analysis and Design Guidelines</a> -
<strong><em>Domain Engineering: Product Lines, Languages, and Conceptual
Models</em></strong>, 2013. [<a
href="https://scholar.google.com/scholar?cluster=15620404537599157753">All
Versions</a>]. In recent years, the development of domain-specific
modeling languages has gained remarkable attention. This is for good
reasons. A domain-specific modeling language incorporates concepts that
represent domain-level knowledge. Hence, systems analysts are not forced
to reconstruct these concepts from scratch. At the same time,
domain-specific modeling languages contribute to model integrity,
because they include already constraints that would otherwise have to be
added manually. Even though there has been a considerable amount of
research on developing and using domain-specific modeling languages,
there is still lack of comprehensive methods to guide the design of
these languages. With respect to the complexity and risk related to
developing a domain-specific modeling language, this is a serious
shortfall. This chapter is aimed at a contribution to filling the gap.
At first, it presents guidelines for selecting a metamodeling language.
Its main focus is on supporting the process from analyzing requirements
to specifying and evaluating a domain-specific modeling
language.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="design-practises">Design Practises</h4>
<ul>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0164121214002799">Quantifying
usability of domain-specific languages: An empirical study on software
maintenance</a> - <strong><em>Journal of Systems and
Software</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=3450893039446010260">All
Versions</a>]. A DSL aims to support software development by offering
abstractions to a particular domain. It is expected that DSLs improve
the maintainability of artifacts otherwise produced with general-purpose
languages. However, the maintainability of the DSL artifacts and, hence,
their adoption in mainstream development, is largely dependent on the
usability of the language itself. Unfortunately, it is often hard to
identify their usability strengths and weaknesses early, as there is no
guidance on how to objectively reveal them. Usability is a multi-faceted
quality characteristic, which is challenging to quantify beforehand by
DSL stakeholders. There is even less support on how to quantitatively
evaluate the usability of DSLs used in maintenance tasks. In this
context, this paper reports a study to compare the usability of textual
DSLs under the perspective of software maintenance. A usability
measurement framework was developed based on the cognitive dimensions of
notations. The framework was evaluated both qualitatively and
quantitatively using two DSLs in the context of two evolving
object-oriented systems. The results suggested that the proposed metrics
were useful: (1) to early identify DSL usability limitations, (2) to
reveal specific DSL features favoring maintenance tasks, and (3) to
successfully analyze eight critical DSL usability dimensions.</p></li>
<li><p><a href="https://dl.acm.org/doi/abs/10.1145/2685028">A Taxonomy
of Domain-Specific Aspect Languages</a> - <strong><em>ACM Computing
Surveys</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=17254174131160041640">All
Versions</a>]. Domain-Specific Aspect Languages (DSALs) are
Domain-Specific Languages (DSLs) designed to express crosscutting
concerns. Compared to DSLs, their aspectual nature greatly amplifies the
language design space. This survey structures this space in order to
shed light on and compare the different domain-specific approaches to
deal with crosscutting concerns. This survey reports on a corpus of 36
DSALs covering the space, discuss a set of design considerations, and
provide a taxonomy of DSAL implementation approaches. This work serves
as a frame of reference to DSAL and DSL researchers, enabling further
advances in the field, and to developers as a guide for DSAL
implementations.</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/abstract/document/9904438">No Grammar
to Rule Them All: A Survey of JSON-style DSLs for Visualization</a> -
<strong><em>IEEE Transactions on Visualization and Computer
Graphics</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=17206818917381447796">All
Versions</a>]. There has been substantial growth in the use of
JSON-based grammars, as well as other standard data serialization
languages, to create visualizations. Each of these grammars serves a
purpose: some focus on particular computational tasks (such as
animation), some are concerned with certain chart types (such as maps),
and some target specific data domains (such as ML). Despite the
prominence of this interface form, there has been little detailed
analysis of the characteristics of these languages. This study surveys
and analyzes the design and implementation of 57 JSON-style DSLs for
visualization. The authors analyze these languages supported by a
collected corpus of examples for each DSL (consisting of 4395 instances)
across a variety of axes organized into concerns related to domain,
conceptual model, language relationships, affordances, and general
practicalities. The authors identify tensions throughout these areas,
such as between formal and colloquial specifications, among types of
users, and within the composition of languages. Through this work, the
authors seek to support language implementers by elucidating the
choices, opportunities, and tradeoffs in visualization DSL
design.</p></li>
<li><p><a href="https://dl.acm.org/doi/abs/10.1145/3622851">How Domain
Experts Use an Embedded DSL</a> - <strong><em>OOPSLA’23</em></strong>,
2023. [<a
href="https://scholar.google.com/scholar?cluster=8416124186663074528">All
Versions</a>]. Programming tools are increasingly integral to research
and analysis in myriad domains, including specialized areas with no
formal relation to computer science. Embedded domain-specific languages
(eDSLs) have the potential to serve these programmers while placing
relatively light implementation burdens on language designers. However,
barriers to eDSL use reduce their practical value and adoption. This
work aims to deepen the understanding of how programmers use eDSLs and
identify user needs to inform future eDSL designs. The authors performed
a contextual inquiry (9 participants) with domain experts using Mimi, an
eDSL for climate change economics modeling. A thematic analysis
identified five key themes, including: the interaction between the eDSL
and the host language has significant and sometimes unexpected impacts
on eDSL user experience, and users preferentially engage with
domain-specific communities and code templates rather than host language
resources.</p></li>
<li><p><a
href="https://link.springer.com/chapter/10.1007/978-981-96-0780-8_9">Abstract
Hardware Grounding Towards the Automated Design of Automation
Systems</a> - <strong><em>ICIRA’24</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=3331524500088540378">All
Versions</a>]. [<a
href="https://arxiv.org/abs/2410.05663">Preprint</a>]. Crafting
automation systems tailored for specific domains requires aligning the
space of human experts’ semantics with the space of robot executable
actions, and scheduling the required resources and system layout
accordingly. Regrettably, there are three major gaps, fine-grained
domain-specific knowledge injection, heterogeneity between human
knowledge and robot instructions, and diversity of users’ preferences,
resulting automation system design a case-by-case and labour-intensive
effort, thus hindering the democratization of automation. This work
refers to this challenging alignment as the abstract hardware grounding
problem, where the authors firstly regard the procedural operations in
humans’ semantics space as the abstraction of hardware requirements,
then the authors ground such abstractions to instantiated hardware
devices, subject to constraints and preferences in the real
world—optimizing this problem is essentially standardizing and
automating the design of automation systems. On this basis, this work
develops an automated design framework in a hybrid data-driven and
principle-derived fashion. Results on designing self-driving
laboratories for enhancing experiment-driven scientific discovery
suggest the proposed framework’s potential to produce compact systems
that fully satisfy domain-specific and user-customized requirements with
no redundancy.</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/abstract/document/10766486">Constraint
Representation Towards Precise Data-Driven Storytelling</a> -
<strong><em>VIS-Gen4DS’24</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=12234019078719898658">All
Versions</a>]. [<a
href="https://arxiv.org/abs/2410.07535">Preprint</a>]. A position paper
on DSL for data-driven storytelling. Data-driven storytelling serves as
a crucial bridge for communicating ideas in a persuasive way. However,
the manual creation of data stories is a multifaceted, labor-intensive,
and case-specific effort, limiting their broader application. As a
result, automating the creation of data stories has emerged as a
significant research thrust. Despite advances in Artificial
Intelligence, the systematic generation of data stories remains
challenging due to their hybrid nature: they must frame a perspective
based on a seed idea in a top-down manner, similar to traditional
storytelling, while coherently grounding insights of given evidence in a
bottom-up fashion, akin to data analysis. These dual requirements
necessitate precise constraints on the permissible space of a data
story. This viewpoint proposes integrating constraints into the data
story generation process. Defined upon the hierarchies of interpretation
and articulation, constraints shape both narrations and illustrations to
align with seed ideas and contextualized evidence. The authors identify
the taxonomy and required functionalities of these constraints. Although
constraints can be heterogeneous and latent, this position paper
explores the potential to represent them in a computation-friendly
fashion via Domain-Specific Languages. The authors believe that
leveraging constraints will facilitate both artistic and scientific
aspects of data story generation.</p></li>
<li><p><a
href="https://www.nature.com/articles/s44160-024-00649-8">Reproducibility
in automated chemistry laboratories using computer science
abstractions</a> - <strong><em>Nature Synthesis</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=2583939834455194329">All
Versions</a>]. While abstraction is critical for the transferability of
automated laboratory science in (bio)chemical and materials sciences,
its improper implementation is a technical debt taken against the
reproducibility of experimental results. Over the decades, computer
science has developed guidelines and strategies for how abstractions are
captured in programming languages—particularly concerning the
substitutability of implementations of abstracted ideas and the clear
definition of the contexts in which abstractions are used. However, few
programming languages developed for automated experiments fully leverage
the wisdom learned in computer science. To achieve collaborative sharing
of scientific knowledge via automated laboratories, the way that
experimental protocols are codified and interpreted by machine agents
must use abstractions responsibly and with reproducibility, rather than
solely transferability, at its core. This Review discusses how computer
science principles of abstraction can be translated to create more
reproducible automation as an enabler for the acceleration of
collaborative research with self-driving laboratories.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="design-automation">Design Automation</h4>
<ul>
<li><p><a href="https://aclanthology.org/2024.acl-long.659/">AutoDSL:
Automated domain-specific language design for structural representation
of procedures with constraints</a> - <strong><em>ACL’24</em></strong>,
2024. [<a
href="https://scholar.google.com/scholar?cluster=588082932830954126">All
Versions</a>]. [<a
href="https://arxiv.org/abs/2406.12324">Preprint</a>]. [<a
href="https://autodsl.org/procedure/papers/acl24shi.html">Project</a>].
The original paper on the automated design of DSLs, referred to as
AutoDSL. Accurate representation of procedures in restricted scenarios,
such as non-standardized scientific experiments, requires precise
depiction of constraints. Unfortunately, Domain-Specific Language (DSL),
as an effective tool to express constraints structurally, often requires
case-by-case hand-crafting, necessitating customized, labor-intensive
efforts. To overcome this challenge, this paper introduces the AutoDSL
framework to automate DSL-based constraint design across various
domains. Utilizing domain specified experimental protocol corpora,
AutoDSL optimizes syntactic constraints and abstracts semantic
constraints. Quantitative and qualitative analyses of the DSLs designed
by AutoDSL across five distinct domains highlight its potential as an
auxiliary module for language models, aiming to improve procedural
planning and execution.</p></li>
<li><p><a
href="https://openreview.net/forum?id=9nUBh4V6SA">Hierarchically
Encapsulated Representation for Protocol Design in Self-Driving Labs</a>
- <strong><em>ICLR’25</em></strong>, 2025. [<a
href="https://scholar.google.com/scholar?cluster=6090102857833092474">All
Versions</a>]. [<a
href="https://autodsl.org/procedure/papers/iclr25shi.html">Project</a>].
Self-driving laboratories have begun to replace human experimenters in
performing single experimental skills or predetermined experimental
protocols. However, as the pace of idea iteration in scientific research
has been intensified by Artificial Intelligence, the demand for rapid
design of new protocols for new discoveries become evident. Efforts to
automate protocol design have been initiated, but the capabilities of
knowledge-based machine designers, such as Large Language Models, have
not been fully elicited, probably for the absence of a systematic
representation of experimental knowledge, as opposed to isolated,
flatten pieces of information. To tackle this issue, this work proposes
a multi-faceted, multi-scale representation, where instance actions,
generalized operations, and product flow models are hierarchically
encapsulated using Domain-Specific Languages. The authors further
develop a data-driven algorithm based on non-parametric modeling that
autonomously customizes these representations for specific domains. The
proposed representation is equipped with various machine designers to
manage protocol design tasks, including planning, modification, and
adjustment. The results demonstrate that the proposed method could
effectively complement Large Language Models in the protocol design
process, serving as an auxiliary module in the realm of machine-assisted
scientific exploration.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="imperative-dsl-applications">Imperative DSL Applications</h4>
<ul>
<li><p><a
href="https://www.science.org/doi/full/10.1126/science.aav2211">Organic
synthesis in a modular robotic system driven by a chemical programming
language</a> - <strong><em>Science</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=13920677955690815682">All
Versions</a>]. [<a
href="https://www.chem.gla.ac.uk/cronin/images/pubs/387-Steiner-ScienceJan19.full.pdf">Preprint</a>].
[<a
href="https://www.science.org/doi/10.1126/science.aav8816">Perspective:
Democratizing synthesis by automation</a>]. This paper develops an
autonomous compiler and robotic laboratory platform to synthesize
organic compounds on the basis of standardized methods descriptions. The
platform comprises conventional equipment such as round-bottom flasks,
separatory funnels, and a rotary evaporator to maximize its
compatibility with extant literature. The authors showcase the system
with short syntheses of three common pharmaceuticals that proceeded
comparably to manual synthesis.</p></li>
<li><p><a
href="https://jbioleng.biomedcentral.com/track/pdf/10.1186/1754-1611-4-13.pdf">Biocoder:
A programming language for standardizing and automating biology
protocols</a> - <strong><em>Journal of Biological
Engineering</em></strong>, 2010. [<a
href="https://scholar.google.com/scholar?start=0&amp;hl=en&amp;as_sdt=0,5&amp;cluster=15572197190838916795">All
Versions</a>]. [<a
href="https://github.com/nmz787/BioCoder">Project</a>]. [<a
href="https://www.microsoft.com/en-us/download/details.aspx?id=52556">Microsoft
Page</a>] This paper introduces BioCoder, a C++ library that enables
biologists to express the exact steps needed to execute a protocol. In
addition to being suitable for automation, BioCoder converts the code
into a readable, English-language description for use by
biologists.</p></li>
<li><p><a
href="https://www.nature.com/articles/s44160-023-00473-6">Universal
chemical programming language for robotic synthesis repeatability</a> -
<strong><em>Nature Synthesis</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=3455106495990439366">All
Versions</a>]. [<a
href="https://www.chem.gla.ac.uk/cronin/images/pubs/rauschen-natsynthesisjan24.pdf">Preprint</a>].
This paper presents an approach that uses a universal chemical
programming language (χDL) to encode and execute synthesis procedures
for a variety of chemical reactions, including reductive amination, ring
formation, esterification, carbon–carbon bond formation and amide
coupling on four different hardware systems in two laboratories. With
around 50 lines of code per reaction, the approach uses abstraction to
efficiently compress chemical protocols.</p></li>
<li><p><a href="https://dl.acm.org/doi/full/10.1145/3604568">Building an
Open Representation for Biological Protocols</a> - <strong><em>ACM
Journal on Emerging Technologies in Computing Systems</em></strong>,
2023. [<a
href="https://scholar.google.com/scholar?cluster=17225405546647782000">All
Versions</a>]. Laboratory protocols are critical to biological research
and development, yet difficult to communicate and reproduce across
projects, investigators, and organizations. While many attempts have
been made to address this challenge, there is currently no available
protocol representation that is unambiguous enough for precise
interpretation and automation, yet simultaneously “human friendly” and
abstract enough to enable reuse and adaptation. The Laboratory Open
Protocol language (LabOP) is a free and open protocol representation
aiming to address this gap, building on a foundation of UML,
Autoprotocol, Aquarium, SBOL RDF, and the Provenance Ontology. LabOP
provides a linked-data representation both for protocols and for records
of their execution and the resulting data, as well as a framework for
exporting from LabOP for execution by either humans or laboratory
automation. LabOP is currently implemented in the form of an RDF
knowledge representation, specification document, and Python library,
and supports execution as manual “paper protocols,” by Autoprotocol or
by Opentrons. From this initial implementation, LabOP is being further
developed as an open community effort.</p></li>
<li><p><a
href="https://dl.acm.org/doi/abs/10.1145/3586183.3606789">KnitScript: A
Domain-Specific Scripting Language for Advanced Machine Knitting</a> -
<strong><em>UIST’23</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?hl=en&amp;as_sdt=0%2C5&amp;q=KnitScript%3A+A+Domain-Specific+Scripting+Language+for+Advanced+Machine+Knitting&amp;btnG=">All
Versions</a>]. [<a
href="https://pypi.org/project/knit-script/">Project</a>]. This paper
presents KnitScript, a domain-specific machine knitting scripting
language that supports computationally driven knitting designs.
KnitScript provides a comprehensive virtual model of knitting machines,
giving access to machine-level capabilities as they are needed while
automating a variety of tedious and error-prone details.</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007/s11119-020-09770-y">A
domain‑specifc language framework for farm management information
systems in precision agriculture</a> - <strong><em>Precision
Agriculture</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=1495954486695213496">All
Versions</a>]. This paper proposes a domain-specific language framework
for the design and development of precision-agriculture FMISs, which
copes with challenges on supporting the understandability, enhancing
communication and analysis of the design decisions, and the
communication among stakeholders.</p></li>
<li><p><a
href="https://openaccess.thecvf.com/content/CVPR2023/html/Raistrick_Infinite_Photorealistic_Worlds_Using_Procedural_Generation_CVPR_2023_paper.html">Infinite
Photorealistic Worlds Using Procedural Generation</a> -
<strong><em>CVPR’23</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=11620922717915489091">All
Versions</a>]. [<a href="https://infinigen.org/">Website</a>]. [<a
href="https://openaccess.thecvf.com/content/CVPR2023/supplemental/Raistrick_Infinite_Photorealistic_Worlds_CVPR_2023_supplemental.pdf">Supplementary
Text</a>]. This paper introduces Infinigen, a procedural generator of
photorealistic 3D scenes of the natural world. Infinigen is entirely
procedural: every asset, from shape to texture, is generated from
scratch via randomized mathematical rules, using no external source and
allowing infinite variation and composition.</p></li>
<li><p><a
href="https://openaccess.thecvf.com/content/CVPR2024/html/Raistrick_Infinigen_Indoors_Photorealistic_Indoor_Scenes_using_Procedural_Generation_CVPR_2024_paper.html">Infinigen
Indoors: Photorealistic Indoor Scenes using Procedural Generation</a> -
<strong><em>CVPR’24</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=14526967027465419958">All
Versions</a>]. This work introduces Infinigen Indoors, a Blender-based
procedural generator of photorealistic indoor scenes. It builds upon the
existing Infinigen system, which focuses on natural scenes, but expands
its coverage to indoor scenes by introducing a diverse library of
procedural indoor assets, including furniture, architecture elements,
appliances, and other day-to-day objects. It also introduces a
constraint-based arrangement system, which consists of a domain-specific
language for expressing diverse constraints on scene composition, and a
solver that generates scene compositions that maximally satisfy the
constraints. The authors provide an export tool that allows the
generated 3D objects and scenes to be directly used for training
embodied agents in real-time simulators such as Omniverse and Unreal.
Infinigen Indoors is open-sourced under the BSD license.</p></li>
<li><p><a href="https://fse.studenttheses.ub.rug.nl/25731/">Corel: A DSL
for Cooking Recipes</a> - 2021. [<a
href="https://scholar.google.com/scholar?cluster=9477049800574267813">All
Versions</a>]. [<a href="https://roorda.dev/recipes/0">Corel recipe
page</a>]. [<a
href="https://www.fao.org/infoods/infoods/tables-and-databases/faoinfoods-databases/en/">International
Network of Food Data Systems (INFOODS)</a>]. The Corel DSL for cooking
recipes enables understanding of and computation with ingredients, and
can construct a nutrition label for the recipe.</p></li>
<li><p><a href="https://dl.acm.org/doi/full/10.1145/3613905.3650756">“We
Need Structured Output”: Towards User-centered Constraints on Large
Language Model Output</a> - <strong><em>CHI EA’24</em></strong>, 2024.
[<a
href="https://scholar.google.com/scholar?cluster=12105435542197416648">All
Versions</a>]. [<a
href="https://research.google/pubs/we-need-structured-output-towards-user-centered-constraints-on-large-language-model-output/">Preprint</a>].
Large language models can produce creative and diverse responses.
However, to integrate them into current developer workflows, it is
essential to constrain their outputs to follow specific formats or
standards. This work surveyed 51 experienced industry professionals to
understand the range of scenarios and motivations driving the need for
output constraints from a user-centered perspective. The authors
identified 134 concrete use cases for constraints at two levels:
low-level, which ensures the output adhere to a structured format and an
appropriate length, and high-level, which requires the output to follow
semantic and stylistic guidelines without hallucination. Critically,
applying output constraints could not only streamline the currently
repetitive process of developing, testing, and integrating LLM prompts
for developers, but also enhance the user experience of LLM-powered
features and applications. The authors conclude with a discussion on
user preferences and needs towards articulating intended constraints for
LLMs, alongside an initial design for a constraint prototyping
tool.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="declarative-dsl-applications">Declarative DSL Applications</h4>
<ul>
<li><p><a href="https://www.nature.com/articles/nbt.1666">The BioPAX
community standard for pathway data sharing</a> - <strong><em>Nature
Biotechnology</em></strong>, 2010. [<a
href="https://scholar.google.com/scholar?cluster=11368332679628594895">All
Versions</a>]. [<a
href="https://core.ac.uk/download/pdf/216139091.pdf">Preprint</a>].
Biological Pathway Exchange (BioPAX) is a standard language to represent
biological pathways at the molecular and cellular level and to
facilitate the exchange of pathway data. BioPAX can represent metabolic
and signaling pathways, molecular and genetic interactions and gene
regulation networks.</p></li>
<li><p><a
href="https://www.science.org/doi/full/10.1126/science.abd7331">Learning
the language of viral evolution and escape</a> -
<strong><em>Science</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=13862653184613223515">All
Versions</a>]. The ability for viruses to mutate and evade the human
immune system and cause infection, called viral escape, remains an
obstacle to antiviral and vaccine development. Understanding the complex
rules that govern escape could inform therapeutic design. This work
modeled viral escape with machine learning algorithms originally
developed for human natural language. The authors identified escape
mutations as those that preserve viral infectivity but cause a virus to
look different to the immune system, akin to word changes that preserve
a sentence’s grammaticality but change its meaning. With this approach,
language models of influenza hemagglutinin, HIV-1 envelope glycoprotein
(HIV Env), and severe acute respiratory syndrome coronavirus 2
(SARS-CoV-2) Spike viral proteins can accurately predict structural
escape patterns using sequence data alone. This study represents a
promising conceptual bridge between natural language and viral
evolution.</p></li>
<li><p><a
href="https://www.biorxiv.org/content/10.1101/2022.12.21.521526v1">A
high-level programming language for generative protein design</a> -
2022. [<a
href="https://scholar.google.com/scholar?cluster=11732741354610784314">All
Versions</a>]. Combining a basic set of building blocks into more
complex forms is a universal design principle. Most protein designs have
proceeded from a manual bottom-up approach using parts created by
nature, but top-down design of proteins is fundamentally hard due to
biological complexity. This work demonstrates how the modularity and
programmability long sought for protein design can be realized through
generative artificial intelligence. Advanced protein language models
demonstrate emergent learning of atomic resolution structure and protein
design principles. The authors leverage these developments to enable the
programmable design of de novo protein sequences and structures of high
complexity. First, the authors describe a high-level programming
language based on modular building blocks that allows a designer to
easily compose a set of desired properties. The authors then develop an
energy-based generative model, built on atomic resolution structure
prediction with a language model, that realizes all-atom structure
designs that have the programmed properties. Designing a diverse set of
specifications, including constraints on atomic coordinates, secondary
structure, symmetry, and multimerization, demonstrates the generality
and controllability of the approach. Enumerating constraints at
increasing levels of hierarchical complexity shows that the approach can
access a combinatorially large design space.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41467-023-39396-3">Artificial
intelligence driven design of catalysts and materials for ring opening
polymerization using a domain-specific language</a> - <strong><em>Nature
Communications</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=6595955912508683146">All
Versions</a>]. [<a
href="https://github.com/IBM/ibm-materials-notebook">Project</a>].
Advances in machine learning (ML) and automated experimentation are
poised to vastly accelerate research in polymer science. Data
representation is a critical aspect for enabling ML integration in
research workflows, yet many data models impose significant rigidity
making it difficult to accommodate a broad array of experiment and data
types found in polymer science. This inflexibility presents a
significant barrier for researchers to leverage their historical data in
ML development. This work shows that a domain specific language, termed
Chemical Markdown Language (CMDL), provides flexible, extensible, and
consistent representation of disparate experiment types and polymer
structures. CMDL enables seamless use of historical experimental data to
fine-tune regression transformer (RT) models for generative molecular
design tasks. The authors demonstrate the utility of this approach
through the generation and the experimental validation of catalysts and
polymers in the context of ring-opening polymerization—although the
authors provide examples of how CMDL can be more broadly applied to
other polymer classes. Critically, this work shows how the CMDL tuned
model preserves key functional groups within the polymer structure,
allowing for experimental validation. These results reveal the
versatility of CMDL and how it facilitates translation of historical
data into meaningful predictive and generative models to produce
experimentally actionable output.</p></li>
<li><p><a href="https://docs.openlaw.io/">OpenLaw</a> -
<strong><em>OpenLaw.io</em></strong>. It is now possible to model all or
parts of legal agreements using code (smart contracts), decreasing the
cost and friction of creating, securing, and generating binding legal
agreements. Lawyers lack basic tools to build these dynamic, “smart”
contracts in a way that is enforceable and understandable to a legal
professional. OpenLaw is a technology stack to help power next
generation “smart” legal agreements, with a domain-specific markup
language, a integration framework, and a series of general
applications.</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007/s10994-021-06120-5">Scenic:
a language for scenario specification and data generation</a> -
<strong><em>Machine Learning</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=13790565080942515865">All
Versions</a>]. This paper proposes a domain-specific language, Scenic,
for describing scenarios that are distributions over scenes and the
behaviors of their agents over time. Scenic combines concise, readable
syntax for spatiotemporal relationships with the ability to
declaratively impose hard and soft constraints over the
scenario.</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/abstract/document/9169399">Domain
Specific Language for Smart Contract Development</a> -
<strong><em>ICBC’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=16998538751745390273">All
Versions</a>]. [<a
href="http://eprints-dev5.cs.univie.ac.at/6341/1/PID6382125.pdf">Preprint</a>].
This research addresses the understanding hardness raised from the
conceptual discrepancy between contractual clauses and corresponding
code of the Solidity programming language, by the design and study of a
domain-specific smart contract language based on higher level of
abstraction that can be automatically transformed to an
implementation.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0950584921002081">iContractML
2.0: A domain-specific language for modeling and deploying smart
contracts onto multiple blockchain platforms</a> -
<strong><em>Information and Software Technology</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=1548144959305241494">All
Versions</a>]. Smart contracts play a vital role in many fields. Despite
being called smart, the development of smart contracts is a tedious task
beyond defining a set of contractual rules. In addition to business
knowledge, coding a smart contract requires strong technical knowledge
in a multiplex of new and rapidly changing domain-specific languages and
blockchain platforms. The goal of this paper is to assist developers in
building smart contracts independently from the language or the target
blockchain platform. In which, this paper presents the second-generation
smart contract language iContractML 2.0. iContractML 2.0 is an
extensible framework that empowers developers to model and generate
functional smart contract code that can be deployed onto multiple
blockchain platforms.</p></li>
<li><p><a href="https://proceedings.mlr.press/v130/lew21a.html">PClean:
Bayesian Data Cleaning at Scale with Domain-Specific Probabilistic
Programming</a> - <strong><em>ICML’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=2892523061439714130">All
Versions</a>]. This work presents PClean, a probabilistic programming
language (PPL) for leveraging dataset-specific knowledge to automate
Bayesian cleaning, automating Bayesian approaches given the diversity of
real-world error patterns and the hardness of inference.</p></li>
<li><p><a href="http://proceedings.mlr.press/v139/tavares21a.html">A
Language for Counterfactual Generative Models</a> -
<strong><em>ICML’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=2067748786482591497">All
Versions</a>]. [<a
href="https://github.com/zenna/Omega.jl">Project</a>]. This paper
presents Omega, a probabilistic programming language with support for
counterfactual inference. This feature is accomplished by introducing a
new operator to probabilistic programming akin to Pearl’s do.</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/abstract/document/6030048">Product
Line Engineering Using Domain-Specific Languages</a> -
<strong><em>ISPLC’11</em></strong>, 2011. [<a
href="https://scholar.google.com/scholar?cluster=17589685299346185442">All
Versions</a>]. [<a
href="https://voelter.de/data/pub/VoelterVisser-PLEusingDSLs.pdf">Preprint</a>].
This paper investigates the application of domain-specific languages in
product line engineering (PLE). It starts by analyzing the limits of
expressivity of feature models. Feature models correspond to
context-free grammars without recursion, which prevents the expression
of multiple instances and references. The authors then show how
domain-specific languages (DSLs) can serve as a middle ground between
feature modeling and programming. They can be used in cases where
feature models are too limited, while keeping the separation between
problem space and solution space provided by feature models. This work
then categorizes useful combinations between configuration with feature
model and construction with DSLs and provide an integration of DSLs into
the conceptual framework of PLE. Finally the authors show how use of a
consistent, unified formalism for models, code, and configuration can
yield important benefits for managing variability and trace
ability.</p></li>
<li><p><a href="https://ieeexplore.ieee.org/document/9613674">A
Domain-Specific Language for Product-Process-Resource Modeling</a> -
<strong><em>ETFA’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=6006131184799036515">All
Versions</a>]. This paper presents the design of the PPR-DSL to
effectively and efficiently represent Product-Process-Resource (PPR)
aspects and evaluate constraints defined for modeling PPR views in the
Formalized Process Description standard (VDI 3682).</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007/s11263-018-1103-5">Configurable
3D Scene Synthesis and 2D Image Rendering with Per-pixel Ground Truth
Using Stochastic Grammars</a> - <strong><em>International Journal of
Computer Vision</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=8301697457354598778">All
Versions</a>]. [<a
href="https://yzhu.io/publication/scenesynthesis2018ijcv/paper.pdf">Preprint</a>].
This work proposes a systematic learning-based approach to the
generation of massive quantities of synthetic 3D scenes and arbitrary
numbers of photorealistic 2D images thereof, with associated ground
truth information, for the purposes of training, benchmarking, and
diagnosing learning-based computer vision and robotics algorithms. In
particular, the authors devise a learning-based pipeline of algorithms
capable of automatically generating and rendering a potentially infinite
variety of indoor scenes by using a stochastic grammar, represented as
an attributed Spatial And-Or Graph, in conjunction with state-of-the-art
physics-based rendering. The pipeline is capable of synthesizing scene
layouts with high diversity, and it is configurable inasmuch as it
enables the precise customization and control of important attributes of
the generated scenes. It renders photorealistic RGB images of the
generated scenes while automatically synthesizing detailed, per-pixel
ground truth data, including visible surface depth and normal, object
identity, and material information (detailed to object parts), as well
as environments (e.g., illuminations and camera viewpoints). The authors
demonstrate the value of the synthesized dataset, by improving
performance in certain machine-learning-based scene understanding
tasks—depth and surface normal prediction, semantic segmentation,
reconstruction, etc.—and by providing benchmarks for and diagnostics of
trained models by modifying object attributes and scene properties in a
controllable manner.</p></li>
<li><p><a href="https://arxiv.org/abs/2410.16770">The Scene Language:
Representing Scenes with Programs, Words, and Embeddings</a> -
<strong><em>CVPR’25</em></strong>, 2025. [<a
href="https://scholar.google.com/scholar?cluster=8704845413716059914">All
Versions</a>]. [<a
href="https://ai.stanford.edu/~yzzhang/projects/scene-language/">Project</a>].
This paper introduces the Scene Language, a visual scene representation
that concisely and precisely describes the structure, semantics, and
identity of visual scenes. It represents a scene with three key
components: a program that specifies the hierarchical and relational
structure of entities in the scene, words in natural language that
summarize the semantic class of each entity, and embeddings that capture
the visual identity of each entity. This representation can be inferred
from pre-trained language models via a training-free inference
technique, given text or image inputs.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="logic-dsl-applications">Logic DSL Applications</h4>
<ul>
<li><p><a
href="https://en.wikipedia.org/wiki/Situation_calculus">Situation
Calculus</a> - <strong><em>Wikipedia</em></strong>. Wikipedia on
Situation Calculus, a logic formalism designed for representing and
reasoning about dynamical domains.</p></li>
<li><p><a
href="https://link.springer.com/content/pdf/10.1007/978-3-030-24658-7.pdf">What
is Answer Set Programming?</a> - <strong><em>Springer</em></strong>,
2008. [<a
href="https://scholar.google.com/scholar?cluster=3691841207891991771">All
Versions</a>]. [<a
href="https://dl.acm.org/doi/abs/10.5555/1620270.1620340">Tutorial on
AAAI</a>]. Answer set programming (ASP) is a form of declarative
programming oriented towards difficult search problems. As an outgrowth
of research on the use of nonmonotonic reasoning in knowledge
representation, it is particularly useful in knowledge-intensive
applications. ASP programs consist of rules that look like Prolog rules,
but the computational mechanisms used in ASP are different: they are
based on the ideas that have led to the creation of fast satisfiability
solvers for propositional logic.</p></li>
<li><p><a
href="https://link.springer.com/chapter/10.1007/3-540-46767-x_28">Answer
Set Programming</a> - <strong><em>ICLPNR’99</em></strong>, 1999. [<a
href="https://scholar.google.com/scholar?cluster=15267370435063454675">All
Versions</a>]. [<a
href="http://people.sabanciuniv.edu/~esraerdem/teaching/krr06/asp.pdf">Preprint</a>].
The original paper on Answer Set Programming (ASP), a form of
declarative programming oriented towards difficult search problems, on
the use of nonmonotonic reasoning in knowledge representation. In ASP
solutions to a problem are represented by answer sets (known also as
stable models), and not by answer substitutions produced in response to
a query, as in conventional logic programming.</p></li>
<li><p><a
href="https://link.springer.com/chapter/10.1007%2F978-3-642-60085-2_16">Action
Languages, Answer Sets, and Planning</a> - <strong><em>The Logic
Programming Paradigms</em></strong>, 1999. [<a
href="https://scholar.google.com/scholar?cluster=2045126541850245645">All
Versions</a>]. [<a
href="https://citeseerx.ist.psu.edu/document?repid=rep1&amp;type=pdf&amp;doi=e58359b3dae3141fd2c85ee3f00c566411134929">Preprint</a>].
This is a discussion of some of the achievements and challenges related
to representing actions and the design of planners from the perspective
of logic programming. The authors talk about recent work on action
languages and translating them into logic programming, on representing
possible histories of an action domain by answer sets, on efficient
implementations of the answer set semantics and their use for generating
plans, and on causal logic and its relation to planning algorithms.
Recent progress in these areas may lead to the creation of planners
which are based on the ideas of logic programming and combine the use of
expressive action description languages with efficient computational
procedures.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/abs/pii/0004370286900731">Qualitative
Simulation</a> - <strong><em>Artificial Intelligence</em></strong>,
1986. [<a
href="https://scholar.google.com/scholar?cluster=4945009733425184345&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://www.cs.utexas.edu/ftp/qsim/papers/Kuipers-aij-86.pdf">Preprint</a>].
This paper presents a precise definition of qualitative structure and
behavior descriptions as abstractions of differential equations and
continuously differentiable functions. The authors present a new
algorithm for qualitative simulation that generalizes the best features
of existing algorithms, and allows direct comparisons among alternate
approaches. Starting with a set of constraints abstracted from a
differential equation, this work proves that the QSIM algorithm is
guaranteed to produce a qualitative behavior corresponding to any
solution to the original equation. The paper also shows that any
qualitative simulation algorithm will sometimes produce spurious
qualitative behaviors: ones which do not correspond to any mechanism
satisfying the given constraints. These observations suggest specific
types of care that must be taken in designing applications of
qualitative causal reasoning systems, and in constructing and validating
a knowledge base of mechanism descriptions.</p></li>
<li><p><a
href="https://www.cs.utexas.edu/users/qr/QR-book.html">Qualitative
Reasoning: Modeling and Simulation with Incomplete Knowledge</a> -
<strong><em>MIT Press</em></strong>, 1994. [<a
href="https://scholar.google.com/scholar?&amp;cluster=6634684154722677465">All
Versions</a>]. This book presents, within a conceptually unified
theoretical framework, a body of methods that have been developed over
the past fifteen years for building and simulating qualitative models of
physical systems - bathtubs, tea kettles, automobiles, the physiology of
the body, chemical processing plants, control systems, electrical
systems - where knowledge of that system is incomplete. The primary tool
for this work is the author’s QSIM algorithm, which is discussed in
detail. Qualitative models are better able than traditional models to
express states of incomplete knowledge about continuous mechanisms.
Qualitative simulation guarantees to find all possible behaviors
consistent with the knowledge in the model. This expressive power and
coverage is important in problem solving for diagnosis, design,
monitoring, explanation, and other applications of artificial
intelligence.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0004370297000507">Qualitative
and quantitative simulation: bridging the gap</a> -
<strong><em>Artificial Intelligence</em></strong>, 1997. [<a
href="https://scholar.google.com/scholar?cluster=9033452473914228535">All
Versions</a>]. Shortcomings of qualitative simulation and of
quantitative simulation motivate combining them to do simulations
exhibiting strengths of both. The resulting class of techniques is
called semiquantitative simulation. One approach to semi-quantitative
simulation is to use numeric intervals to represent incomplete
quantitative information. This research demonstrates semi-quantitative
simulation using intervals in an implemented semi-quantitative simulator
called Q3. Q3 progressively refines a qualitative simulation, providing
increasingly specific quantitative predictions which can converge to a
numerical simulation in the limit while retaining important correctness
guarantees from qualitative and interval simulation techniques.</p></li>
<li><p><a href="https://pubs.acs.org/doi/10.1021/acssynbio.8b00229">A
Logic Programming Language for Computational Nucleic Acid Devices</a> -
<strong><em>ACS Synthetic Biology</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=3336951672389047784">All
Versions</a>]. This paper presents a logic programming language that
allows a broad range of computational nucleic acid systems to be
designed and analyzed. The language extends standard logic programming
with a novel equational theory to express nucleic acid molecular motifs.
It automatically identifies matching motifs present in the full system,
in order to apply a specified transformation expressed as a logical
rule.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41596-021-00675-2">Genetic
circuit design automation with Cello 2.0</a> - <strong><em>Nature
Protocol</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=7418307542591684967">All
Versions</a>]. [<a
href="https://www.researchgate.net/profile/Samuel-Oliveira-38/publication/358801979_Genetic_circuit_design_automation_with_Cello_20/links/635debf412cbac6a3e0b19e4/Genetic-circuit-design-automation-with-Cello-20.pdf">Preprint</a>].
Cells interact with their environment, communicate among themselves,
track time and make decisions through functions controlled by natural
regulatory genetic circuits consisting of interacting biological
components. Synthetic programmable circuits used in therapeutics and
other applications can be automatically designed by computer-aided
tools. The Cello software designs the DNA sequences for programmable
circuits based on a high-level software description and a library of
characterized DNA parts representing Boolean logic gates. This process
allows for design specification reuse, modular DNA part library curation
and formalized circuit transformations based on experimental data. This
protocol describes Cello 2.0, a freely available cross-platform software
written in Java. Cello 2.0 enables flexible descriptions of the logic
gates’ structure and their mathematical models representing dynamic
behavior, new formal rules for describing the placement of gates in a
genome, a new graphical user interface, support for Verilog 2005 syntax
and a connection to the SynBioHub parts repository software environment.
Collectively, these features expand Cello’s capabilities beyond
Escherichia coli plasmids to new organisms and broader genetic contexts,
including the genome. Designing circuits with Cello 2.0 produces an
abstract Boolean network from a Verilog file, assigns biological parts
to each node in the Boolean network, constructs a DNA sequence and
generates highly structured and annotated sequence representations
suitable for downstream processing and fabrication, respectively. The
result is a sequence implementing the specified Boolean function in the
organism and predictions of circuit performance. Depending on the size
of the design space and users’ expertise, jobs may take minutes or hours
to complete.</p></li>
<li><p><a href="https://arxiv.org/abs/2502.13372">MoVer: Motion
Verification for Motion Graphics Animations</a> - <strong><em>ACM
Transactions on Graphics</em></strong>, 2025. [<a
href="https://scholar.google.com/scholar?cluster=527747131334466686">All
Versions</a>]. While large vision-language models can generate motion
graphics animations from text prompts, they regularly fail to include
all of spatio-temporal properties described in the prompt. This work
introduces MoVer, a motion verification DSL based on first-order logic
that can check spatio-temporal properties of a motion graphics
animation. The authors identify a general set of such properties that
people commonly use to describe animations (e.g., the direction and
timing of motions, the relative positioning of objects, etc.). The
authors implement these properties as predicates in MoVer and provide an
execution engine that can apply a MoVer program to any input SVG-based
motion graphics animation. The authors then demonstrate how MoVer can be
used in an LLM-based synthesis and verification pipeline for iteratively
refining motion graphics animations. Given a text prompt, the pipeline
synthesizes a motion graphics animation and a corresponding MoVer
program. Executing the verification program on the animation yields a
report of the predicates that failed and the report can be automatically
fed back to LLM to iteratively correct the animation.</p></li>
<li><p><a href="https://openreview.net/forum?id=C45YqeBDUM">The
KoLMogorov Test: Compression by Code Generation</a> -
<strong><em>ICLR’25</em></strong>, 2025. [<a
href="https://scholar.google.com/scholar?cluster=16809888292456252135">All
Versions</a>]. Compression is at the heart of intelligence. A
theoretically optimal way to compress any sequence of data is to find
the shortest program that outputs that sequence and then halts. However,
such Kolmogorov compression is uncomputable, and code generating LLMs
struggle to approximate this theoretical ideal, as it requires
reasoning, planning and search capabilities beyond those of current
models. This work introduces the KoLMogorov-Test (KT), a
compression-as-intelligence intelligence test for code generation LLMs.
In KT a model is presented with a sequence of data at inference time,
and asked to generate the shortest DSL (designed specifically for the
task) program that produces the sequence. The authors identify several
benefits of KT for both evaluation and training: an essentially infinite
number of problem instances of varying difficulty is readily available,
strong baselines already exist, the evaluation metric (compression)
cannot be gamed, and pretraining data contamination is highly unlikely.
To evaluate current models, the authors use audio, text, and DNA data,
as well as sequences produced by random synthetic DSL programs.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41598-022-21801-4">Meta-analysis
of the functional neuroimaging literature with probabilistic logic
programming</a> - <strong><em>Scientific Reports</em></strong>, 2022.
[<a
href="https://scholar.google.com/scholar?cluster=5952076495542489316">All
Versions</a>]. Inferring reliable brain-behavior associations requires
synthesizing evidence from thousands of functional neuroimaging studies
through meta-analysis. However, existing meta-analysis tools are limited
to investigating simple neuroscience concepts and expressing a
restricted range of questions. This work expands the scope of
neuroimaging meta-analysis by designing NeuroLang: a domain-specific
language to express and test hypotheses using probabilistic first-order
logic programming. By leveraging formalisms found at the crossroads of
artificial intelligence and knowledge representation, NeuroLang provides
the expressivity to address a larger repertoire of hypotheses in a
meta-analysis, while seamlessly modeling the uncertainty inherent to
neuroimaging data. The authors demonstrate the language’s capabilities
in conducting comprehensive neuroimaging meta-analysis through use-case
examples that address questions of structure-function associations.
Specifically, the authors infer the specific functional roles of three
canonical brain networks, support the role of the visual word-form area
in visuospatial attention, and investigate the heterogeneous
organization of the frontoparietal control network.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="dsl-program-synthesis">DSL Program Synthesis</h4>
<ul>
<li><p><a
href="https://dl.acm.org/doi/abs/10.1145/3220134.3220135">pix2code:
Generating Code from a Graphical User Interface Screenshot</a> -
<strong><em>ACM SIGCHI Symposium on Engineering Interactive Computing
Systems</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=8296741513177971931">All
Versions</a>]. [<a
href="https://github.com/tonybeltramelli/pix2code">Code</a>]. [<a
href="https://uizard.io/research/">Website</a>]. This paper shows that
deep learning methods can be leveraged to train a model end-to-end to
automatically reverse engineer user interfaces and generate code from a
single input image with over 77% of accuracy for three different
platforms (i.e. iOS, Android and web-based technologies).</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2018/hash/6788076842014c83cedadbe6b0ba0314-Abstract.html">Learning
to Infer Graphics Programs from Hand-Drawn Images</a> -
<strong><em>NeurIPS’18</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=14065112485794121024">All
Versions</a>]. The method learns a model that uses program synthesis
techniques to recover a graphics program from drawing primitives. These
programs have constructs like variable bindings, iterative loops, or
simple kinds of conditionals. With a graphics program in hand, we can
correct errors made by the deep network and extrapolate
drawings.</p></li>
<li><p><a href="https://dl.acm.org/doi/abs/10.1145/3571207">babble:
Learning Better Abstractions with E-Graphs and Anti-unification</a> -
<strong><em>POPL’23</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=7935064016901049715">All
Versions</a>]. This paper proposes library learning modulo theory
(LLMT), a new library learning algorithm that additionally takes as
input an equational theory for a given problem domain. LLMT uses
e-graphs and equality saturation to compactly represent the space of
programs equivalent modulo the theory, and uses a novel e-graph
anti-unification technique to find common patterns in the corpus more
directly and efficiently.</p></li>
<li><p><a href="https://dl.acm.org/doi/abs/10.1145/3571234">Top-Down
Synthesis for Library Learning</a> - <strong><em>POPL’23</em></strong>,
2023. [<a
href="https://scholar.google.com/scholar?cluster=12324277007659029766">All
Versions</a>]. This paper introduces corpus-guided top-down synthesis as
a mechanism for synthesizing library functions that capture common
functionality from a corpus of programs in a domain specific language
(DSL). The algorithm builds abstractions directly from initial DSL
primitives, using syntactic pattern matching of intermediate
abstractions to intelligently prune the search space and guide the
algorithm towards abstractions that maximally capture shared structures
in the corpus.</p></li>
<li><p><a
href="https://royalsocietypublishing.org/doi/full/10.1098/rsta.2022.0050">DreamCoder:
growing generalizable, interpretable knowledge with wake–sleep Bayesian
program learning</a> - <strong><em>Philosophical Transactions of the
Royal Society A</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=11356436337624711843">All
Versions</a>]. [<a
href="https://arxiv.org/abs/2006.08381">Preprint</a>]. This paper
presents DreamCoder, a system that learns to solve problems by writing
programs. It builds expertise by creating domain-specific programming
languages for expressing domain concepts, together with neural networks
to guide the search for programs within these languages. A ‘wake–sleep’
learning algorithm alternately extends the language with new symbolic
abstractions and trains the neural network on imagined and replayed
problems. DreamCoder solves both classic inductive programming tasks and
creative tasks such as drawing pictures and building scenes.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41467-022-32012-w">Synthesizing
theories of human language with Bayesian program induction</a> -
<strong><em>Nature Communications</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=8603772394100237159">All
Versions</a>]. Automated, data-driven construction and evaluation of
scientific models and theories is a long-standing challenge in
artificial intelligence. This work presents a framework for
algorithmically synthesizing models of a basic part of human language:
morpho-phonology, the system that builds word forms from sounds. The
authors integrate Bayesian inference with program synthesis and
representations inspired by linguistic theory and cognitive models of
learning and discovery. Across 70 datasets from 58 diverse languages,
the system synthesizes human-interpretable models for core aspects of
each language’s morpho-phonology, sometimes approaching models posited
by human linguists. Joint inference across all 70 data sets
automatically synthesizes a meta-model encoding interpretable
cross-language typological tendencies. Finally, the same algorithm
captures few-shot learning dynamics, acquiring new morphophonological
rules from just one or a few examples. These results suggest routes to
more powerful machine-enabled discovery of interpretable models in
linguistics and other scientific domains.</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper_files/paper/2023/hash/cd40d0d65bfebb894ccc9ea822b47fa8-Abstract-Conference.html">Grammar
Prompting for Domain-Specific Language Generation with Large Language
Models</a> - <strong><em>NeurIPS’23</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=11694070042468483715">All
Versions</a>]. Grammar prompting is a simple approach to enable LLMs to
use external knowledge and domain-specific constraints expressed through
a grammar in Backus–Naur Form (BNF) during in-context learning. Grammar
prompting augments each demonstration example with a specialized grammar
that is minimally sufficient for generating the particular output
example, where the specialized grammar is a subset of the full DSL
grammar. For inference, the LLM first predicts a BNF grammar given a
test input, and then generates the output according to the rules of the
grammar. Experiments demonstrate that grammar prompting can enable LLMs
to perform competitively on a diverse set of DSL generation tasks,
including semantic parsing (SMCalFlow, Overnight, GeoQuery), PDDL
planning, and SMILES-based molecule generation.</p></li>
<li><p><a href="https://arxiv.org/abs/2303.14100">Errors are Useful
Prompts: Instruction Guided Task Programming with Verifier-Assisted
Iterative Prompting</a> - 2023. [<a
href="https://scholar.google.com/scholar?cluster=8063693456660536915">All
Versions</a>]. [<a
href="https://github.com/ac-rad/xdl-generation">Project</a>]. [<a
href="https://ac-rad.github.io/clairify/">Website</a>]. This paper
proposes CLAIRIFY, an approach that combines automatic iterative
prompting with program verification to ensure programs written in
data-scarce domain-specific language are syntactically valid and
incorporate environment constraints.</p></li>
<li><p><a
href="https://dl.acm.org/doi/full/10.1145/3613904.3642319">PhotoScout:
Synthesis-Powered Multi-Modal Image Search</a> - <strong><em>ACM
SIGCHI’24</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=6522231014055730719">All
Versions</a>]. This paper explores a new multi-modal image search
approach that allows users to conveniently specify and perform semantic
image search tasks. With the tool, PhotoScout, the user interactively
provides natural language descriptions, positive and negative examples,
and object tags to specify their search tasks. Under the hood,
PhotoScout is powered by a program synthesis engine that generates
visual queries in a domain-specific language and executes the
synthesized program to retrieve the desired images.</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper_files/paper/2024/hash/54dd9e0cff6d9214e20d97eb2a3bae49-Abstract-Conference.html">Expert-level
protocol translation for self-driving labs</a> -
<strong><em>NeurIPS’24</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=13997597682274906943">All
Versions</a>]. [<a
href="https://autodsl.org/procedure/papers/neurips24shi.html">Project</a>].
Recent development in Artificial Intelligence (AI) models has propelled
their application in scientific discovery, but the validation and
exploration of these discoveries require subsequent empirical
experimentation. The concept of self-driving laboratories promises to
automate and thus boost the experimental process following AI-driven
discoveries. However, the transition of experimental protocols,
originally crafted for human comprehension, into formats interpretable
by machines presents significant challenges, which, within the context
of specific expert domain, encompass the necessity for structured as
opposed to natural language, the imperative for explicit rather than
tacit knowledge, and the preservation of causality and consistency
throughout protocol steps. Presently, the task of protocol translation
predominantly requires the manual and labor-intensive involvement of
domain experts and information technology specialists, rendering the
process time-intensive. To address these issues, this work proposes a
framework that automates the protocol translation process through a
three-stage workflow, which incrementally constructs Protocol Dependence
Graphs (PDGs) that approach structured on the syntax level, completed on
the semantics level, and linked on the execution level. Quantitative and
qualitative evaluations have demonstrated its performance at par with
that of human experts, underscoring its potential to significantly
expedite and democratize the process of scientific discovery by
elevating the automation capabilities within self-driving
laboratories.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41586-023-06924-6">Mathematical
discoveries from program search with large language models</a> -
<strong><em>Nature</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=5653439474813913484">All
Versions</a>]. Large language models (LLMs) have demonstrated tremendous
capabilities in solving complex tasks, from quantitative reasoning to
understanding natural language. However, LLMs sometimes suffer from
confabulations (or hallucinations), which can result in them making
plausible but incorrect statements1,2. This hinders the use of current
large models in scientific discovery. This work introduces FunSearch
(short for searching in the function space), an evolutionary procedure
based on pairing a pretrained LLM with a systematic evaluator. The
authors demonstrate the effectiveness of this approach to surpass the
best-known results in important problems, pushing the boundary of
existing LLM-based approaches3. Applying FunSearch to a central problem
in extremal combinatorics—the cap set problem—we discover new
constructions of large cap sets going beyond the best-known ones, both
in finite dimensional and asymptotic cases. This shows that it is
possible to make discoveries for established open problems using LLMs.
The authors showcase the generality of FunSearch by applying it to an
algorithmic problem, online bin packing, finding new heuristics that
improve on widely used baselines. In contrast to most computer search
approaches, FunSearch searches for programs that describe how to solve a
problem, rather than what the solution is. Beyond being an effective and
scalable strategy, discovered programs tend to be more interpretable
than raw solutions, enabling feedback loops between domain experts and
FunSearch, and the deployment of such programs in real-world
applications.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="cognitive-foundations">Cognitive Foundations</h4>
<ul>
<li><p><a
href="https://www.cell.com/trends/cognitive-sciences/fulltext/S1364-6613(20)30174-1">The
Child as Hacker</a> - <strong><em>Trends in Cognitive
Sciences</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=13128656954836679743">All
Versions</a>]. The scope of human learning and development poses a
radical challenge for cognitive science. The authors propose that
developmental theories can address this challenge by adopting
perspectives from computer science. Many of our best models treat
learning as analogous to computer programming because symbolic programs
provide the most compelling account of sophisticated mental
representations. The authors specifically propose that children’s
learning is analogous to a particular style of programming called
hacking, making code better along many dimensions through an open-ended
set of goals and activities. By contrast to existing theories, which
depend primarily on local search and simple metrics, this view
highlights the many features of good mental representations and the
multiple complementary processes children use to create them.</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper_files/paper/2022/hash/182aed0379591ebd1d655b2bdc152075-Abstract-Datasets_and_Benchmarks.html">Communicating
Natural Programs to Humans and Machines</a> -
<strong><em>NeurIPS’22</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=13381039702346039142">All
Versions</a>]. While humans readily generate and interpret instructions
in a general language, computer systems are shackled to a narrow
domain-specific language that they can precisely execute. This makes
building intelligent systems that can generalize to novel situations
such as ARC difficult. Human-generated instructions are referred as
“natural programs”. While they resemble computer programs, they are
distinct in two ways: First, they contain a wide range of primitives;
Second, they frequently leverage communicative strategies beyond
directly executable codes.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41467-024-50966-x">Symbolic
metaprogram search improves learning efficiency and explains rule
learning in humans</a> - <strong><em>Nature
Communications</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=7670274141609367282">All
Versions</a>]. Symbolic models based on program learning successfully
explain rule-learning in many domains, but performance degrades quickly
as program complexity increases. It remains unclear how to scale
symbolic rule-learning methods to model human performance in challenging
domains. This work shows that symbolic search over the space of
metaprograms—programs that revise programs—dramatically improves
learning efficiency. On a behavioral benchmark of 100 algorithmically
rich rules, this approach fits human learning more accurately than
alternative models while also using orders of magnitude less search. The
computation required to match median human performance is consistent
with conservative estimates of human thinking time. The results suggest
that metaprogram-like representations may help human learners to
efficiently acquire rules.</p></li>
<li><p><a href="https://www.nature.com/articles/s41562-025-02227-0">How
laypeople evaluate scientific explanations containing jargon</a> -
<strong><em>Nature Human Behavior</em></strong>, 2025. [<a
href="https://scholar.google.com/scholar?cluster=6467855047925175367">All
Versions</a>]. Individuals rely on others’ expertise to achieve a basic
understanding of the world. But how can non-experts achieve
understanding from explanations that, by definition, they are
ill-equipped to assess? Across 9 experiments with 6,698 participants
(Study 1A = 737; 1B = 734; 1C = 733; 2A = 1,014; 2B = 509; 2C = 1,012;
3A = 1,026; 3B = 512; 4 = 421), this work addresses this puzzle by
focusing on scientific explanations with jargon. The authors identify
‘when’ and ‘why’ the inclusion of jargon makes explanations more
satisfying, despite decreasing their comprehensibility. The authors find
that jargon increases satisfaction because laypeople assume the jargon
fills gaps in explanations that are otherwise incomplete. The authors
also identify strategies for debiasing these judgements: when people
attempt to generate their own explanations, inflated judgements of poor
explanations with jargon are reduced, and people become better
calibrated in their assessments of their own ability to
explain.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="problem-solving">Problem Solving</h3>
<h4 id="human-level-problem-solving">Human-Level Problem Solving</h4>
<ul>
<li><p><a href="https://psycnet.apa.org/record/1959-07883-001">Elements
of a theory of human problem solving</a> - <strong><em>Psychological
Review</em></strong>, 1958. [<a
href="https://scholar.google.com/scholar?cluster=6226995019045187501">All
Versions</a>]. Herbert Simon’s original idea on human problem
solving.</p></li>
<li><p><a href="https://psycnet.apa.org/record/1973-10478-000">Human
Problem Solving</a> - <strong><em>Englewood Cliffs, NJ:
Prentice-hall</em></strong>, 1972. [<a
href="https://scholar.google.com/scholar?cluster=3996229083126262536&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Herbert Simon’s classic idea of human problem solving as
search.</p></li>
<li><p><a
href="http://196.223.158.148/bitstream/handle/123456789/2978/596.pdf?sequence=1&amp;isAllowed=y">Learning
to Solve Problems: A Handbook for Designing Problem-Solving Learning
Environments</a> - <strong><em>Taylorfrancis</em></strong>, 2010. [<a
href="https://scholar.google.com/scholar?cluster=13262690779319271809&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.science.org/doi/abs/10.1126/science.185.4157.1124">Judgment
under Uncertainty: Heuristics and Biases: Biases in judgments reveal
some heuristics of thinking under uncertainty</a> -
<strong><em>Science</em></strong>, 1974. [<a
href="https://scholar.google.com/scholar?cluster=17040257859216791312&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Daniel Kahneman’s classic idea of prospective
theory.</p></li>
<li><p><a
href="https://www.pnas.org/content/117/47/29381.short">Computational
evidence for hierarchically structured reinforcement learning in
humans</a> - <strong><em>Proceedings of the National Academy of
Sciences</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=5731363475904675608&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A piece of evidence on hierarchical human
planning.</p></li>
<li><p><a
href="https://www.cnbc.cmu.edu/braingroup/papers/sarafyazd_jazayeri_2019.pdf">Hierarchical
reasoning by neural circuits in the frontal cortex</a> -
<strong><em>Science</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=9875733886908769773&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Neuroscience evidence supporting rule switch.</p></li>
<li><p><a href="https://oar.princeton.edu/rt4ds/file/11875/2161">The
importance of mixed selectivity in complex cognitive tasks</a> -
<strong><em>Nature</em></strong>, 2013. [<a
href="https://scholar.google.com/scholar?cluster=2707751672275136220&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper introducing mixed selectivity with
high-dimensional neural representations.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41586-022-04743-9">People
construct simplified mental representations to plan</a> -
<strong><em>Nature</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=12068944400080889789&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A computational account on rational problem
representation in human planning.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S1364661322002819">Goals,
usefulness and abstraction in value-based choice</a> -
<strong><em>Trends in Cognitive Sciences</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=6256990098976657651&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>]. A review that outlines the computational and biological
principles that enable the brain to compute the usefulness of an option
or action by creating abstractions that flexibly adapt to changing
goals.</p></li>
<li><p><a href="https://elifesciences.org/articles/68943">Value signals
guide abstraction during learning</a> - <strong><em>eLife</em></strong>,
2022. [<a
href="https://scholar.google.com/scholar?cluster=10324834842795908439&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007/BF00058926">Learning to
perceive and act by trial and error</a> - <strong><em>Machine
Learning</em></strong>, 1991. [<a
href="https://scholar.google.com/scholar?cluster=1987606770603964473&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://onlinelibrary.wiley.com/doi/abs/10.1207/s15516709cog1801_3">Representations
in distributed cognitive tasks</a> - <strong><em>Cognitive
Science</em></strong>, 1994. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=14781266698447195483">All
Versions</a>].</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/abs/pii/S0364021399800226">The
nature of external representations in problem solving</a> -
<strong><em>Cognitive Science</em></strong>, 1997. [<a
href="https://scholar.google.com/scholar?cluster=10698887231200401430&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.pnas.org/content/pnas/117/47/29302.full.pdf">Rapid
trail-and-error learning with simulation supports flexible tool use and
physical reasoning.</a> - <strong><em>Proceedings of the National
Academy of Sciences</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=14400178089019636923">All
Versions</a>]. [<a
href="https://sites.google.com/view/virtualtoolsgame/home">Project</a>].
[<a
href="https://www.pnas.org/content/pnas/suppl/2020/11/20/1912341117.DCSupplemental/pnas.1912341117.sapp.pdf">Appendix</a>].
Many animals, and an increasing number of artificial agents, display
sophisticated capabilities to perceive and manipulate objects. But human
beings remain distinctive in their capacity for flexible, creative tool
use—using objects in new ways to act on the world, achieve a goal, or
solve a problem. To study this type of general physical problem solving,
this work introduces the Virtual Tools game. In this game, people solve
a large range of challenging physical puzzles in just a handful of
attempts. This work proposes that the flexibility of human physical
problem solving rests on an ability to imagine the effects of
hypothesized actions, while the efficiency of human search arises from
rich action priors which are updated via observations of the world. The
authors instantiate these components in the “sample, simulate, update”
(SSUP) model and show that it captures human performance across 30
levels of the Virtual Tools game. More broadly, this model provides a
mechanism for explaining how people condense general physical knowledge
into actionable, task-specific plans to achieve flexible and efficient
physical problem solving.</p></li>
<li><p><a
href="https://cognitivesciencesociety.org/cogsci20/papers/0765/0765.pdf">Abstract
strategy learning underlies flexible transfer in physical problem
solving</a> - <strong><em>CogSci’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?hl=en&amp;as_sdt=0%2C5&amp;q=Abstract+strategy+learning+underlies+flexible+transfer+in+physical+problem+solving.&amp;btnG=">All
Versions</a>].</p></li>
<li><p><a href="https://openreview.net/forum?id=CXyZrKPz4CU">Physion:
Evaluating Physical Prediction from Vision in Humans and Machines</a> -
<strong><em>NeurIPS’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=8733318111076645893&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S2352154620301236">Exploration:
from machines to humans</a> - <strong><em>Current Opinion in Behavioral
Sciences</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=8015078432419172621&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S2352154620301467">Balancing
exploration and exploitation with information and randomization</a> -
<strong><em>Current Opinion in Behavioral Sciences</em></strong>, 2021.
[<a
href="https://scholar.google.com/scholar?cluster=8164388137243077863&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0092867421008369">Hippocampal
neurons construct a map of an abstract value space</a> -
<strong><em>Cell</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=12658820581876003172&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.pnas.org/content/106/25/10370.short">Insightful
problem solving and creative tool modification by captive nontool-using
rooks</a> - <strong><em>Proceedings of the National Academy of
Sciences</em></strong>, 2009. [<a
href="https://scholar.google.com/scholar?cluster=6776471679661065229&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://www.pnas.org/content/suppl/2009/05/28/0901008106.DCSupplemental">Supplementary
Material</a>]. A piece of evidence on creative tool use in intelligent
animals.</p></li>
<li><p><a
href="https://cpilab.org/pubs/Dasgupta2018Learning.pdf">Learning to act
by integrating mental simulations and physical experiments</a> -
<strong><em>CogSci’18</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=7342920174595829739&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://github.com/ishita-dg/SimulationVSAction">Code</a>].</p></li>
<li><p><a href="https://gershmanlab.com/pubs/Momennejad17.pdf">The
successor representation in human reinforcement learning</a> -
<strong><em>Nature Human Behavior</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=7317529612823134939&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.science.org/doi/abs/10.1126/scirobotics.aav3150">Beyond
imitation: Zero-shot task transfer on robots by learning concepts as
cognitive programs</a> - <strong><em>Science Robotics</em></strong>,
2019. [<a
href="https://scholar.google.com/scholar?cluster=7451223471302689228">All
Versions</a>]. Humans can infer concepts from image pairs and apply
those in the physical world in a completely different setting, enabling
tasks like IKEA assembly from diagrams. If robots could represent and
infer high-level concepts, then it would notably improve their ability
to understand our intent and to transfer tasks between different
environments. To that end, the authors introduce a computational
framework that replicates aspects of human concept learning. Concepts
are represented as programs on a computer architecture consisting of a
visual perception system, working memory, and action controller. The
instruction set of this cognitive computer has commands for parsing a
visual scene, directing gaze and attention, imagining new objects,
manipulating the contents of a visual working memory, and controlling
arm movement. Inferring a concept corresponds to inducing a program that
can transform the input to the output. Some concepts require the use of
imagination and recursion. Previously learned concepts simplify the
learning of subsequent, more elaborate concepts and create a hierarchy
of abstractions. The authors demonstrate how a robot can use these
abstractions to interpret novel concepts presented to it as schematic
images and then apply those concepts in very different situations. By
bringing cognitive science ideas on mental imagery, perceptual symbols,
embodied cognition, and deictic mechanisms into the realm of machine
learning, this work brings researchers closer to the goal of building
robots that have interpretable representations and common
sense.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="planning">Planning</h4>
<ul>
<li><p><a href="https://jair.org/index.php/jair/article/view/11175">From
Skills to Symbols: Learning Symbolic Representations for Abstract
High-Level Planning</a> - <strong><em>Journal of Artificial Intelligence
Research</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=17962480659445514879">All
Versions</a>]. This work considers the problem of constructing abstract
representations for planning in high-dimensional, continuous
environments. The authors assume an agent equipped with a collection of
high-level actions, and construct representations provably capable of
evaluating plans composed of sequences of those actions. The authors
first consider the deterministic planning case, and show that the
relevant computation involves set operations performed over sets of
states. The authors then consider probabilistic planning, which they
show requires generalizing from sets of states to distributions over
states. Finally, the authors apply these techniques to create a physical
robot system that autonomously learns its own symbolic representation of
a mobile manipulation task directly from sensorimotor data—point clouds,
map locations, and joint angles—and then plans using that
representation.</p></li>
<li><p><a
href="https://www.annualreviews.org/doi/abs/10.1146/annurev-control-091420-084139">Integrated
Task and Motion Planning</a> - <strong><em>Annual Review of Control,
Robotics, and Autonomous Systems</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=478421650694199529">All
Versions</a>]. The problem of planning for a robot that operates in
environments containing a large number of objects, taking actions to
move itself through the world as well as to change the state of the
objects, is known as task and motion planning (TAMP). TAMP problems
contain elements of discrete task planning, discrete–continuous
mathematical programming, and continuous motion planning and thus cannot
be effectively addressed by any of these fields directly. In this
article, the authors define a class of TAMP problems and survey
algorithms for solving them, characterizing the solution methods in
terms of their strategies for solving the continuous-space subproblems
and their techniques for integrating the discrete and continuous
components of the search.</p></li>
<li><p><a
href="https://dspace.mit.edu/handle/1721.1/126626">Differentiable
Physics and Stable Modes for Tool-Use and Manipulation Planning</a> -
<strong><em>Robotics: Science and Systems</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=10342169019935480143&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://gershmanlab.com/pubs/Dasgupta18_simulation.pdf">Learning
to act by integrating mental simulations and physical experiments</a> -
<strong><em>CogSci’21</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=7342920174595829739&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://onlinelibrary.wiley.com/doi/ftr/10.1111/cogs.12928">What
Is the Model in Model-Based Planning?</a> - <strong><em>Cognitive
Science</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=10598397017491369972&amp;hl=en&amp;scisbd=1&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/2109.11082.pdf">Discovering State
and Action Abstractions for Generalized Task and Motion Planning</a> -
<strong><em>AAAI’22</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=1054368060554971920">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="intrinsic-motivation">Intrinsic Motivation</h4>
<ul>
<li><p><a
href="https://proceedings.neurips.cc/paper/2004/hash/4be5a36cbaca8ab9d2066debfe4e65c1-Abstract.html">Intrinsically
Motivated Reinforcement Learning</a> -
<strong><em>NeurIPS’04</em></strong>, 2004. [<a
href="https://scholar.google.com/scholar?cluster=9736217847061704054&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A comprehensive review on intrinsic reward functions in
classic reinforcement learning.</p></li>
<li><p><a
href="https://www.frontiersin.org/articles/10.3389/neuro.12.006.2007/full">What
is intrinsic motivation? A typology of computational approaches</a> -
<strong><em>Frontiers in Neurorobotics</em></strong>, 2009. [<a
href="https://scholar.google.com/scholar?cluster=11901343819872275353&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.jair.org/index.php/jair/article/view/12087">Adapting
Behavior via Intrinsic Reward: A Survey and Empirical Study</a> -
<strong><em>Journal of Artificial Intelligence Research</em></strong>,
2020. [<a
href="https://scholar.google.com/scholar?cluster=5309595875334344707&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://proceedings.mlr.press/v70/pathak17a.html">Curiosity-driven
Exploration by Self-supervised Prediction</a> -
<strong><em>ICML’17</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=9379743003299559904&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper on curiosity as intrinsic
motivation.</p></li>
<li><p><a href="https://arxiv.org/abs/1706.01502">UCB Exploration via
Q-Ensembles</a> - 2017. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=13260404166621290240">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/2010.03110">Causal Curiosity: RL
Agents Discovering Self-supervised Experiments for Causal Representation
Learning</a> - <strong><em>ICML’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=4880520597219138666&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2015/hash/e00406144c1e7e35240afed70f34166a-Abstract.html">Variational
Information Maximisation for Intrinsically Motivated Reinforcement
Learning</a> - <strong><em>NeurIPS’15</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=9262504233068870193&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper on empowerment as intrinsic
motivation.</p></li>
<li><p><a href="https://psyarxiv.com/ybs7g/">Intrinsic Exploration as
Empowerment in a Richly Structured Online Game</a> - 2022. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=12321757821600526668">All
Versions</a>].</p></li>
<li><p><a href="https://gershmanlab.com/pubs/Tomov21.pdf">Multi-task
reinforcement learning in humans</a> - <strong><em>Nature Human
Behavior</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=14589018692074515644&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/abstract/document/10778628">JARVIS-1:
Open-World Multi-Task Agents With Memory-Augmented Multimodal Language
Models</a> - <strong><em>IEEE Transactions on Pattern Analysis and
Machine Intelligence</em></strong>. [<a
href="https://scholar.google.com/scholar?cluster=12845806504666245406">All
Versions</a>]. Achieving human-like planning and control with multimodal
observations in an open world is a key milestone for more functional
generalist agents. Existing approaches can handle certain long-horizon
tasks in an open world. However, they still struggle when the number of
open-world tasks could potentially be infinite and lack the capability
to progressively enhance task completion as game time progresses. This
work introduces JARVIS-1, an open-world agent that can perceive
multimodal input (visual observations and human instructions), generate
sophisticated plans, and perform embodied control, all within the
popular yet challenging open-world Minecraft universe. Specifically, the
authors develop JARVIS-1 on top of pre-trained multimodal language
models, which map visual observations and textual instructions to plans.
The plans will be ultimately dispatched to the goal-conditioned
controllers. JARVIS-1 is outfitted with a multimodal memory, which
facilitates planning using both pre-trained knowledge and its actual
game survival experiences. JARVIS-1 is the existing most general agent
in Minecraft, capable of completing over 200 different tasks using
control and observation space similar to humans.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="reinforcement-learning">Reinforcement Learning</h4>
<ul>
<li><p><a
href="https://www.andrew.cmu.edu/user/rmorina/papers/SuttonBook.pdf">Reinforcement
learning: An introduction</a> - <strong><em>MIT Press</em></strong>,
2018. [<a
href="https://scholar.google.com/scholar?cluster=8821915215029978039&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Richard Sutton’s comprehensive book on reinforcement
learning.</p></li>
<li><p><a
href="https://www.jair.org/index.php/jair/article/view/10166">Reinforcement
learning: A survey</a> - <strong><em>Journal of Artificial Intelligence
Research</em></strong>, 1996. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=4983604491168613713">All
Versions</a>]. Leslie Kaelbling’s review on reinforcement
learning.</p></li>
<li><p><a href="https://arxiv.org/pdf/2011.00583.pdf">An overview of
multi-agent reinforcement learning from game theoretical perspective</a>
- 2020. [<a
href="https://scholar.google.com/scholar?cluster=16197919002723407603&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Yaodong Yang’s review on multi-agent reinforcement
learning from the perspective of game theory.</p></li>
<li><p><a
href="https://klab.tch.harvard.edu/academia/classes/Neuro230/ReadingAssignments/MnihEtAlHassibis15NatureControlDeepRL.pdf">Human-level
control through deep reinforcement learning</a> -
<strong><em>Nature</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=12439121588427761338&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper on solving Atari games via Deep
Q-Network.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0004370299000521">Between
MDPs and semi-MDPs: A framework for temporal abstraction in
reinforcement learning</a> - <strong><em>Artificial
Intelligence</em></strong>, 1999. [<a
href="https://scholar.google.com/scholar?cluster=1471968208408231068&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper on operation reinforcement
learning.</p></li>
<li><p><a
href="http://oucsace.cs.ohio.edu/~chelberg/classes/680/paperPresentations/NathanPaperToPresent.pdf">On
Monte Carlo Tree Search and Reinforcement Learning</a> -
<strong><em>Journal of Artificial Intelligence Research</em></strong>,
2017. [<a
href="https://scholar.google.com/scholar?cluster=5805718077259491860&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/1805.00909">Reinforcement Learning
and Control as Probabilistic Inference: Tutorial and Review</a> - 2018.
[<a
href="https://scholar.google.com/scholar?cluster=16437288987337534404&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="http://rail.eecs.berkeley.edu/deeprlcourse-fa18/static/slides/lec-15.pdf">Slides</a>].
Sergey Levine’s tutorial on treating reinforcement learning
probabilisticly.</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2019/hash/4a46fbfca3f1465a27b210f4bdfe6ab3-Abstract.html">A
Generalized Algorithm for Multi-Objective Reinforcement Learning and
Policy Adaptation</a> - <strong><em>NeurIPS’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=7721047641895252765&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://openreview.net/forum?id=9SS69KwomAM">Solving
Compositional Reinforcement Learning Problems via Task Reduction</a> -
<strong><em>ICLR’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=15628616147808752058&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&amp;arnumber=8460689">Neural
Task Programming: Learning to Generalize Across Hierarchical Tasks</a> -
<strong><em>ICRA’18</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=7155333517647976638&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://academic.oup.com/logcom/article-abstract/28/2/337/4695480">Learning
to act: qualitative learning of deterministic action models</a> -
<strong><em>Journal of Logic and Computation</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=14570482854600886953&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/2109.06076">Learning to Act and
Observe in Partially Observable Domains</a> - 2021. [<a
href="https://scholar.google.com/scholar?cluster=2258600434630687063&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/2107.06277">Why Generalization in
RL is Difficult: Epistemic POMDPs and Implicit Partial Observability</a>
- <strong><em>NeurIPS’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=9640851185758072663&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A formal treatment on the generalization problem in
reinforcement learning.</p></li>
<li><p><a href="https://openreview.net/forum?id=r1nTpv9eg">Learning to
Perform Physics Experiments via Deep Reinforcement Learning</a> -
<strong><em>ICLR’17</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=13142558595749186250&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/abstract/document/9387127">Data-Efficient
Learning for Complex and Real-Time Physical Problem Solving Using
Augmented Simulation</a> - <strong><em>Robotics and Automation
Letters</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=3140653562829320759&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.jmlr.org/papers/volume18/16-634/16-634.pdf">A Survey
of Preference-Based Reinforcement Learning Methods</a> -
<strong><em>Journal of Machine Learning Research</em></strong>, 2017.
[<a
href="https://scholar.google.com/scholar?cluster=13278778479251450967&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://papers.NeurIPS.cc/paper/2021/file/4079016d940210b4ae9ae7d41c4a2065-Paper.pdf">On
the Expressivity of Markov Reward</a> -
<strong><em>NeurIPS’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=4524686816939437211&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A formal treatment of tasks and rewards in reinforcement
learning modeling.</p></li>
<li><p><a href="https://proceedings.mlr.press/v37/schulman15.html">Trust
Region Policy Optimization</a> - <strong><em>ICML’15</em></strong>,
2015. [<a
href="https://scholar.google.com/scholar?cluster=4215501129336400677&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper introducing TRPO, a method for
optimizing control policies, with guaranteed monotonic
improvement.</p></li>
<li><p><a
href="http://proceedings.mlr.press/v70/achiam17a/achiam17a.pdf">Constrained
Policy Optimization</a> - <strong><em>ICML’17</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=6114366704163518185&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper on constrained reinforcement learning
(safe reinforcement learning).</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper_files/paper/2019/hash/5faf461eff3099671ad63c6f3f094f7f-Abstract.html">When
to Trust Your Model: Model-Based Policy Optimization</a> -
<strong><em>NeurIPS’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=4248859125840907707&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://bair.berkeley.edu/blog/2019/12/12/mbpo/">Post</a>].</p></li>
<li><p><a href="http://proceedings.mlr.press/v139/lee21g.html">SUNRISE:
A Simple Unified Framework for Ensemble Learning in Deep Reinforcement
Learning</a> - <strong><em>ICML’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=8840831494454574191&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://github.com/pokaxpoka/sunrise">Code</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/2202.13252">The Quest for a Common
Model of the Intelligent Decision Maker</a> -
<strong><em>Multi-disciplinary Conference on Reinforcement Learning and
Decision Making’22</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=7652784232757502910&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Richard Sutton’s perspective on the future directions of
reinforcement learning research.</p></li>
<li><p><a
href="https://dl.acm.org/doi/abs/10.5555/3491440.3492111">Automatic
curriculum learning for deep RL: a short survey</a> -
<strong><em>IJCAI’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=10660055557098312214&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://proceedings.mlr.press/v139/romac21a.html">TeachMyAgent: a
Benchmark for Automatic Curriculum Learning in Deep RL</a> -
<strong><em>ICML’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=11016662361926634008&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://github.com/flowersteam/TeachMyAgent">Project</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="inverse-reinforcement-learning">Inverse Reinforcement
Learning</h4>
<ul>
<li><p><a
href="https://dl.acm.org/doi/pdf/10.1145/1015330.1015430">Apprenticeship
Learning via Inverse Reinforcement Learning</a> -
<strong><em>ICML’04</em></strong>, 2004. [<a
href="https://scholar.google.com/scholar?cluster=10260011060619377707&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Pieter Abbeel and Andrew Ng’s original paper on inverse
reinforcement learning (IRL).</p></li>
<li><p><a
href="https://www.ijcai.org/Proceedings/07/Papers/416.pdf">Bayesian
Inverse Reinforcement Learning</a> - <strong><em>IJCAI’07</em></strong>,
2007. [<a
href="https://scholar.google.com/scholar?cluster=4154724070362583557&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A Bayesian account on classic inverse reinforcement
learning.</p></li>
<li><p><a href="https://arxiv.org/abs/1902.07742">From Language to
Goals: Inverse Reinforcement Learning for Vision-Based Instruction
Following</a> - <strong><em>ICLR’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=9128320307925997063&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/1904.06317.pdf">Few-shot Bayesian
imitation learning with logical program policies.</a> -
<strong><em>AAAI’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=5103854692762145813&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="http://export.arxiv.org/pdf/2011.09854">Generalized
Inverse Planning: Learning Lifted non-Markovian Utility for
Generalizable Task Representation</a> - 2020. [<a
href="https://scholar.google.com/scholar?cluster=18369106870663956780&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://proceedings.mlr.press/v139/malik21a.html">Inverse
Constrained Reinforcement Learning</a> -
<strong><em>ICML’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?hl=en&amp;as_sdt=0%2C5&amp;q=Inverse+Constrained+Reinforcement+Learning+S+Malik&amp;btnG=">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="system-1-system-2">System 1 &amp; System 2</h3>
<h4 id="dual-coding-theory">Dual-Coding Theory</h4>
<ul>
<li><p><a href="https://zh.pb1lib.org/book/1004349/825277">Mental
Representations: A Dual Coding Approach</a> - <strong><em>Oxford
University Press</em></strong>, 1990. [<a
href="https://scholar.google.com/scholar?hl=en&amp;as_sdt=0,5&amp;q=mental+representations:+a+dual+coding+approach">All
Versions</a>]. The original book on dual coding theory, in the
neuroscience account of mental representation.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S1364661321001765">Dual
coding of knowledge in the human brain</a> - <strong><em>Trends in
Cognitive Sciences</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=11751507203561842501&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Yanchao Bi’s review on neuroscience experiments on dual
coding theory.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0896627320302798">Two
Forms of Knowledge Representations in the Human Brain</a> -
<strong><em>Neuron</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=16941965185680116049">All
Versions</a>]. Illustrating language-derived and sensory-derived
knowledge.</p></li>
<li><p><a
href="http://bilab.bnu.edu.cn/paper/2018/Wang_2018_Cerebral_Cortex.pdf">Organizational
Principles of Abstract Words in the Human Brain</a> -
<strong><em>Cerebral Cortex</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=15272192531353715481">All
Versions</a>].</p></li>
<li><p><a
href="http://bilab.bnu.edu.cn/paper/2022/Fu_2022_CC.pdf">Different
computational relations in language are captured by distinct brain
systems</a> - <strong><em>Cerebral Cortex</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=720215181903530260&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://europepmc.org/article/med/28190038">The
Deese-Roediger-McDermott (DRM) task: A simple cognitive paradigm to
investigate false memories in the laboratory</a> - <strong><em>Journal
of Visualized Experiments</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=10880194606861797581&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://mri-q.com/uploads/3/4/5/7/34572113/gallant_piis0896627312009348.pdf">A
continuous semantic space describes the representation of thousands of
object and action categories across the human brain</a> -
<strong><em>Neuron</em></strong>, 2012. [<a
href="https://scholar.google.com/scholar?cluster=10348115268396987731&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.nature.com/articles/s41562-021-01259-6">Rational
arbitration between statistics and rules in human sequence
processing</a> - <strong><em>Nature Human Behavior</em></strong>, 2022.
[<a
href="https://scholar.google.com/scholar?cluster=9856085207409198966&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="neural-symbolic-ai">Neural-Symbolic AI</h4>
<ul>
<li><p><a
href="https://link.springer.com/chapter/10.1007/978-3-642-59789-3_58">Regression
Analysis for Interval-Valued Data</a> - <strong><em>Data Analysis,
Classification, and Related Methods</em></strong>, 2000. [<a
href="https://scholar.google.com/scholar?cluster=9407097855380377791&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper on symbolic regression.</p></li>
<li><p><a
href="https://link.springer.com/chapter/10.1007/978-3-7908-1709-6_20">Symbolic
data analysis: what is it?</a> - <strong><em>Proceedings in
Computational Statistics</em></strong>, 2006. [<a
href="https://scholar.google.com/scholar?cluster=3730437602749399283&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/1805.10872">DeepProbLog: Neural
Probabilistic Logic Programming</a> -
<strong><em>NeurIPS’18</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=6079567413300944995&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper on neuro-symbolic probabilistic
programming.</p></li>
<li><p><a
href="https://www.jair.org/index.php/jair/article/view/11172">Learning
Explanatory Rules from Noisy Data</a> - <strong><em>Journal of
Artificial Intelligence Research</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=2553893814364678772&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper for differential Inductive Logic
Programming.</p></li>
<li><p><a
href="https://cs.nju.edu.cn/zhouzh/zhouzh.files/publication/aaai17lasin.pdf">Combining
Logical Abduction and Statistical Induction: Discovering Written
Primitives with Human Knowledge</a> - <strong><em>AAAI’17</em></strong>,
2017. [<a
href="https://scholar.google.com/scholar?cluster=14477085725208589393&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/1904.10729.pdf">Neural Logic
Reinforcement Learning</a> - <strong><em>ICML’19</em></strong>, 2019.
[<a
href="https://scholar.google.com/scholar?cluster=18074632043038701502&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://papers.NeurIPS.cc/paper/8548-bridging-machine-learning-and-logical-reasoning-by-abductive-learning">Bridging
Machine Learning and Logical Reasoning by Abductive Learning.</a> -
<strong><em>NeurIPS’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=1518342375288126288&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://daiwz.net/org/slides/ABL-meetup.html#/slide-title">Slides</a>].
[<a href="https://github.com/AbductiveLearning/ABL-HED">Code</a>]. The
original paper on Abductive Learning, a derivative-free approach for
neuro-symbolic learning.</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007/s11432-018-9801-4">Abductive
learning: towards bridging machine learning and logical reasoning</a> -
<strong><em>Science China Information Sciences</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=8541635351775190855&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/2010.03514.pdf">Abductive
Knowledge Induction From Raw Data</a> -
<strong><em>IJCAI’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=7027142960863064076&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2021/hash/df7e148cabfd9b608090fa5ee3348bfe-Abstract.html">Fast
Abductive Learning by Similarity-based Consistency Optimization</a> -
<strong><em>NeurIPS’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=8539963460239876225&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. An approach for accelerating the convergence of Abductive
Learning.</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2019/file/c20a7ce2a627ba838cfbff082db35197-Paper.pdf">Learning
by Abstraction: The Neural State Machine</a> -
<strong><em>NeurIPS’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=7361406080192630148&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0004370220301855">Making
sense of sensory input</a> - <strong><em>Artificial
Intelligence</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=11875529139573472578&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/2103.14230v1.pdf">Abstract
Spatial-Temporal Reasoning via Probabilistic Abduction and Execution</a>
- <strong><em>CVPR’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=4172146500538799638&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://openreview.net/pdf?id=SJlh8CEYDB">Learn to
explain efficiently via neural logic inductive learning</a> -
<strong><em>ICLR’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=4550874980727321525&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://github.com/gblackout/NLIL">Project</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/2006.06649">Closed Loop
Neural-Symbolic Learning via Integrating Neural Perception, Grammar
Parsing, and Symbolic Reasoning</a> - <strong><em>ICML’20</em></strong>,
2020. [<a
href="https://scholar.google.com/scholar?cluster=9257372000778020812&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/2003.08978">Generating new
concepts with hybrid neuro-symbolic models.</a> -
<strong><em>CogSci’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=1912020791698331044">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/2006.14448">Learning Task-General
Representations with Generative Neuro-Symbolic Modeling</a> -
<strong><em>ICLR’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=1335404082385789329">All
Versions</a>].</p></li>
<li><p><a
href="http://clgiles.ist.psu.edu/IST597/materials/slides/papers-memory/2016-graves.pdf">Hybrid
computing using a neural network with dynamic external memory</a> -
<strong><em>Nature</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=8100274942961380405&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.science.org/doi/full/10.1126/sciadv.aay2631">AI
Feynman: A physics-inspired method for symbolic regression</a> -
<strong><em>Science Advances</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=3655502646441210453">All
Versions</a>]. A core challenge for both physics and artificial
intelligence (AI) is symbolic regression: finding a symbolic expression
that matches data from an unknown function. Although this problem is
likely to be NP-hard in principle, functions of practical interest often
exhibit symmetries, separability, compositionality, and other
simplifying properties. In this spirit, the authors develop a recursive
multidimensional symbolic regression algorithm that combines neural
network fitting with a suite of physics-inspired techniques. The authors
apply it to 100 equations from the Feynman Lectures on Physics, and it
discovers all of them, while previous publicly available software cracks
only 71; for a more difficult physics-based test set, this work improves
the state-of-the-art success rate from 15 to 90%.</p></li>
<li><p><a
href="http://papers.NeurIPS.cc/paper/8546-classification-by-components-probabilistic-modeling-of-reasoning-over-a-set-of-components.pdf">Classification-by-Components:
Probabilistic Modeling of Reasoning over a Set of Components</a> -
<strong><em>NeurIPS’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=12691103404451941071&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/2006.11524.pdf">Neuro-Symbolic
Visual Reasoning: Disentangling “Visual” from “Reasoning”</a> -
<strong><em>ICML’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=13160160974887139307&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2020/file/0d82627e10660af39ea7eb69c3568955-Paper.pdf">Understanding
Deep Architectures with Reasoning Layer</a> -
<strong><em>NeurIPS’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=937882599430270789&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/1905.10307.pdf">An Explicitly
Relational Neural Network Architecture</a> -
<strong><em>ICML’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=37732747764322837&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/2103.01937.pdf">Neural Production
Systems</a> - <strong><em>ICML’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=15299280949648915581&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Yoshua Bengio’s perspective on slot attention model as a
general production system.</p></li>
<li><p><a href="https://arxiv.org/pdf/2008.06662.pdf">Compositional
Generalization via Neural-Symbolic Stack Machines</a> -
<strong><em>NeurIPS’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=15612498612943317331&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://openreview.net/pdf?id=H1eSS3CcKX">Stochastic
Optimization of Sorting Networks via Continuous Relaxations</a> -
<strong><em>ICLR’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=10619362619006891050&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://openreview.net/pdf?id=BkxUvnEYDH">Program Guided
Agent</a> - <strong><em>ICLR’20</em></strong>, 2020. [<a
href="https://openreview.net/forum?id=BkxUvnEYDH">All
Versions</a>].</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2020/hash/7a685d9edd95508471a9d3d6fcace432-Abstract.html">Learning
Compositional Rules via Neural Program Synthesis</a> -
<strong><em>NeurIPS’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=3160670555314650508&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/2006.11287">Discovering Symbolic
Models from Deep Learning with Inductive Biases</a> -
<strong><em>NeurIPS’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=9452091824686227240&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/1904.11694.pdf">Neural Logic
Machines</a> - <strong><em>ICLR’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=4525183211642569463&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/1904.12584.pdf">The Neuro-Symbolic
Concept Learner: Interpreting Scenes, Words, and Sentences From Natural
Supervision</a> - <strong><em>ICLR’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=8837128214653317831&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://papers.NeurIPS.cc/paper/2019/file/98d8a23fd60826a2a474c5b4f5811707-Paper.pdf">Visual
Concept-Metaconcept Learning</a> - <strong><em>NeurIPS’19</em></strong>,
2019. [<a
href="https://scholar.google.com/scholar?cluster=1888051343232298875&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/2103.16564">Grounding Physical
Concepts of Objects and Events Through Dynamic Visual Reasoning</a> -
<strong><em>ICLR’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=16735976343684307244&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://jiajunwu.com/papers/toqnet_ijcai.pdf">Temporal
and Object Quantification Networks</a> -
<strong><em>IJCAI’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=17251222943638414124&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/2009.01719.pdf">Grounded Language
Learning Fast and Slow</a> - <strong><em>ICLR’21</em></strong>, 2021.
[<a
href="https://scholar.google.com/scholar?cluster=17735027444431750346&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://github.com/deepmind/dm_fast_mapping?s=05">Project</a>].</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007/s10994-022-06142-7">Detect,
Understand, Act: A Neuro-symbolic Hierarchical Reinforcement Learning
Framework</a> - <strong><em>Machine Learning</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=10321228117236432485&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A neuro-symbolic framework that integrates meta-policy
learning in inductive logic programming.</p></li>
<li><p><a
href="https://openaccess.thecvf.com/content/CVPR2023/html/Gupta_Visual_Programming_Compositional_Visual_Reasoning_Without_Training_CVPR_2023_paper.html">Visual
Programming: Compositional Visual Reasoning Without Training</a> -
<strong><em>CVPR’23</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=16156060658942400125&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. VISPROG, a neuro-symbolic approach to solving complex and
compositional visual tasks given natural language instructions, using
the in-context learning ability of large language models to generate
python-like modular programs, which are then executed to get both the
solution and a comprehensive and interpretable rationale.</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/abstract/document/9338352">Semi-Supervised
Abductive Learning and Its Application to Theft Judicial Sentencing</a>
- <strong><em>ICDM’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=16646246740380524224">All
Versions</a>]. [<a
href="https://www.lamda.nju.edu.cn/huangyx/src/ICDM20-SSABL.pdf">Preprint</a>].
In many practical tasks, there are usually two kinds of common
information: cheap unlabeled data and domain knowledge in the form of
symbols. There are some attempts using one single information source,
such as semi-supervised learning and abductive learning. However, there
is little work to use these two kinds of information sources at the same
time, because it is very difficult to combine symbolic logical
representation and numerical model optimization effectively. The
learning becomes even more challenging when the domain knowledge is
insufficient. This paper presents an attempt-Semi-Supervised ABductive
Learning (SS-ABL) framework. In this framework, semi-supervised learning
is trained via pseudo labels of unlabeled data generated by abductive
learning, and the background knowledge is refined via the label
distribution predicted by semi-supervised learning. The above framework
can be optimized iteratively and can be naturally interpretable. The
effectiveness of the framework has been fully verified in the theft
judicial sentencing of real legal documents. In the case of missing
sentencing elements and mixed legal rules, the framework is apparently
superior to many existing baseline practices, and provides explanatory
assistance to judicial sentencing.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="explainability">Explainability</h3>
<h4 id="trustworthy-ai">Trustworthy AI</h4>
<ul>
<li><p><a
href="https://www.pnas.org/doi/full/10.1073/pnas.2111547119">Bayesian
modeling of human–AI complementarity</a> - <strong><em>Proceedings of
the National Academy of Sciences</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=15735143859968841009&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A Bayesian framework for combining the predictions and
different types of confidence scores from humans and machines.</p></li>
<li><p><a
href="https://www.science.org/doi/10.1126/scirobotics.aay4663">A tale of
two explanations: Enhancing human trust by explaining robot behavior</a>
- <strong><em>Science Robotics</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=3985046411399524590">All
Versions</a>]. [<a
href="https://yzhu.io/publication/openbottle2019scirob/paper.pdf">Preprint</a>].
The ability to provide comprehensive explanations of chosen actions is a
hallmark of intelligence. Lack of this ability impedes the general
acceptance of AI and robot systems in critical tasks. This paper
examines what forms of explanations best foster human trust in machines
and proposes a framework in which explanations are generated from both
functional and mechanistic perspectives. The robot system learns from
human demonstrations to open medicine bottles using (i) an embodied
haptic prediction model to extract knowledge from sensory feedback, (ii)
a stochastic grammar model induced to capture the compositional
structure of a multistep task, and (iii) an improved Earley parsing
algorithm to jointly leverage both the haptic and grammar models. The
robot system not only shows the ability to learn from human
demonstrators but also succeeds in opening new, unseen bottles. Using
different forms of explanations generated by the robot system, we
conducted a psychological experiment to examine what forms of
explanations best foster human trust in the robot. The authors found
that comprehensive and real-time visualizations of the robot’s internal
decisions were more effective in promoting human trust than explanations
based on summary text descriptions. In addition, forms of explanation
that are best suited to foster trust do not necessarily correspond to
the model components contributing to the best task performance. This
divergence shows a need for the robotics community to integrate model
components to enhance both task execution and human trust in
machines.</p></li>
<li><p><a href="https://arxiv.org/abs/1909.06907">X-ToM: Explaining with
Theory-of-Mind for Gaining Justified Human Trust</a> - <strong><em>CVPR
XAI Workshop’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=7751326666821697923">All
Versions</a>]. This work presents a new explainable AI (XAI) framework
aimed at increasing justified human trust and reliance in the AI machine
through explanations. The authors pose explanation as an iterative
communication process, i.e. dialog, between the machine and human user.
More concretely, the machine generates sequence of explanations in a
dialog which takes into account three important aspects at each dialog
turn: (a) human’s intention (or curiosity); (b) human’s understanding of
the machine; and (c) machine’s understanding of the human user. To do
this, the authors use Theory of Mind (ToM) which helps us in explicitly
modeling human’s intention, machine’s mind as inferred by the human as
well as human’s mind as inferred by the machine. In other words, these
explicit mental representations in ToM are incorporated to learn an
optimal explanation policy that takes into account human’s perception
and beliefs. Furthermore, the authors also show that ToM facilitates in
quantitatively measuring justified human trust in the machine by
comparing all the three mental representations. We applied our framework
to three visual recognition tasks, namely, image classification, action
recognition, and human body pose estimation. The authors argue that our
ToM based explanations are practical and more natural for both expert
and non-expert users to understand the internal workings of complex
machine learning models. This is the first work to derive explanations
using ToM. Extensive human study experiments verify our hypotheses,
showing that the proposed explanations significantly outperform the
state-of-the-art XAI methods in terms of all the standard quantitative
and qualitative XAI evaluation metrics including human trust, reliance,
and explanation satisfaction.</p></li>
<li><p><a
href="https://ojs.aaai.org/index.php/AAAI/article/view/5643">CoCoX:
Generating Conceptual and Counterfactual Explanations via
Fault-Lines</a> - <strong><em>AAAI’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=17443137068166403183&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S2589004221015510">CX-ToM:
Counterfactual explanations with theory-of-mind for enhancing human
trust in image recognition models</a> -
<strong><em>iScience</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=17526041764295337444">All
Versions</a>]. This work proposes CX-ToM, short for counterfactual
explanations with theory-of-mind, a new explainable AI (XAI) framework
for explaining decisions made by a deep convolutional neural network
(CNN). In contrast to the current methods in XAI that generate
explanations as a single shot response, the authors pose explanation as
an iterative communication process, i.e., dialogue between the machine
and human user. More concretely, this CX-ToM framework generates a
sequence of explanations in a dialogue by mediating the differences
between the minds of the machine and human user. To do this, the authors
use Theory of Mind (ToM) which helps us in explicitly modeling the
human’s intention, the machine’s mind as inferred by the human, as well
as human’s mind as inferred by the machine. Moreover, most
state-of-the-art XAI frameworks provide attention (or heat map) based
explanations. In this work, the authors show that these attention-based
explanations are not sufficient for increasing human trust in the
underlying CNN model. In CX-ToM, the authors instead use counterfactual
explanations called fault-lines which are defined as follows: given an
input image I for which a CNN classification model M predicts class
cpred, a fault-line identifies the minimal semantic-level features
(e.g., stripes on zebra), referred to as explainable concepts, that need
to be added to or deleted from I to alter the classification category of
I by M to another specified class calt. Extensive experiments verify the
hypotheses, demonstrating that CX-ToM significantly outperforms the
state-of-the-art XAI models.</p></li>
<li><p><a
href="https://www.nature.com/articles/s42256-023-00692-8">Explaining
machine learning models with interactive natural language conversations
using TalkToModel</a> - <strong><em>Nature Machine
Intelligence</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=7044008493489695982">All
Versions</a>]. Practitioners increasingly use machine learning (ML)
models, yet models have become more complex and harder to understand. To
understand complex models, researchers have proposed techniques to
explain model predictions. However, practitioners struggle to use
explainability methods because they do not know which explanation to
choose and how to interpret the explanation. This work addresses the
challenge of using explainability methods by proposing TalkToModel: an
interactive dialogue system that explains ML models through natural
language conversations. TalkToModel consists of three components: an
adaptive dialogue engine that interprets natural language and generates
meaningful responses; an execution component that constructs the
explanations used in the conversation; and a conversational interface.
In real-world evaluations, 73% of healthcare workers agreed they would
use TalkToModel over existing systems for understanding a disease
prediction model, and 85% of ML professionals agreed TalkToModel was
easier to use, demonstrating that TalkToModel is highly effective for
model explainability.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="strong-machine-learning">Strong Machine Learning</h4>
<ul>
<li><p><a
href="https://link.springer.com/article/10.1007/s10994-018-5707-3">Ultra-Strong
Machine Learning: comprehensibility of programs learned with ILP</a> -
<strong><em>Machine Learning</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=17551060457946144913">All
Versions</a>]. During the 1980s Michie defined Machine Learning in terms
of two orthogonal axes of performance: predictive accuracy and
comprehensibility of generated hypotheses. Since predictive accuracy was
readily measurable and comprehensibility not so, later definitions in
the 1990s, such as Mitchell’s, tended to use a one-dimensional approach
to Machine Learning based solely on predictive accuracy, ultimately
favouring statistical over symbolic Machine Learning approaches. In this
paper the authors provide a definition of comprehensibility of
hypotheses which can be estimated using human participant trials. The
authors present two sets of experiments testing human comprehensibility
of logic programs. In the first experiment we test human
comprehensibility with and without predicate invention. Results indicate
comprehensibility is affected not only by the complexity of the
presented program but also by the existence of anonymous predicate
symbols. In the second experiment the authors directly test whether any
state-of-the-art ILP systems are ultra-strong learners in Michie’s
sense, and select the Metagol system for use in humans trials. Results
show participants were not able to learn the relational concept on their
own from a set of examples but they were able to apply the relational
definition provided by the ILP system correctly. This implies the
existence of a class of relational concepts which are hard to acquire
for humans, though easy to understand given an abstract explanation. The
authors believe improved understanding of this class could have
potential relevance to contexts involving human learning, teaching and
verbal interaction.</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007%2Fs10994-020-05941-0">Beneficial
and harmful explanatory machine learning</a> - <strong><em>Machine
Learning</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=16983722694047294963">All
Versions</a>]. Given the recent successes of Deep Learning in AI there
has been increased interest in the role and need for explanations in
machine learned theories. A distinct notion in this context is that of
Michie’s definition of ultra-strong machine learning (USML). USML is
demonstrated by a measurable increase in human performance of a task
following provision to the human of a symbolic machine learned theory
for task performance. A recent paper demonstrates the beneficial effect
of a machine learned logic theory for a classification task, yet no
existing work has examined the potential harmfulness of machine’s
involvement for human comprehension during learning. This paper
investigates the explanatory effects of a machine learned theory in the
context of simple two person games and proposes a framework for
identifying the harmfulness of machine explanations based on the
Cognitive Science literature. The approach involves a cognitive window
consisting of two quantifiable bounds and it is supported by empirical
evidence collected from human trials. The quantitative and qualitative
results indicate that human learning aided by a symbolic machine learned
theory which satisfies a cognitive window has achieved significantly
higher performance than human self learning. Results also demonstrate
that human learning aided by a symbolic machine learned theory that
fails to satisfy this window leads to significantly worse performance
than unaided human learning.</p></li>
<li><p><a href="https://www.ijcai.org/Proceedings/2017/497">Deep Forest:
Towards An Alternative to Deep Neural Networks</a> -
<strong><em>IJCAI’17</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=7391596872731517007">All
Versions</a>]. [<a
href="https://github.com/LAMDA-NJU/Deep-Forest">Project</a>]. This paper
proposes gcForest, a decision tree ensemble approach with performance
highly competitive to deep neural networks in a broad range of tasks. In
contrast to deep neural networks which require great effort in
hyper-parameter tuning, gcForest is much easier to train; even when it
is applied to different data across different domains in the
experiments, excellent performance can be achieved by almost same
settings of hyper-parameters. The training process of gcForest is
efficient, and users can control training cost according to
computational resource available. The efficiency may be further enhanced
because gcForest is naturally apt to parallel implementation.
Furthermore, in contrast to deep neural networks which require
large-scale training data, gcForest can work well even when there are
only small-scale training data.</p></li>
<li><p><a href="https://arxiv.org/abs/2004.00221">NBDT: Neural-Backed
Decision Trees</a> - <strong><em>NeurIPS’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=1902399007162005819&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://github.com/alvinwan/neural-backed-decision-trees">Code</a>].
Expliciting the decision process of a decision tree through neural
networks.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="explainable-deep-learning">Explainable Deep Learning</h4>
<ul>
<li><p><a
href="https://github.com/jacobgil/pytorch-grad-cam">pytorch-grad-cam</a>
- 2021. Class Activation Map methods implemented in Pytorch, with many
elegant features.</p></li>
<li><p><a href="https://ieeexplore.ieee.org/document/8099837">Network
dissection: Quantifying interpretability of deep visual
representations</a> - <strong><em>CVPR’17</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=18069685615852396783&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a href="http://netdissect.csail.mit.edu/">Project</a>].
[<a href="http://places2.csail.mit.edu/index.html">Dataset:
Places365</a>]. The original paper on visualizing the class activation
maps to explain convolutional neural networks.</p></li>
<li><p><a
href="https://www.pnas.org/content/pnas/early/2020/08/31/1907375117.full.pdf">Understanding
the role of Individual Units in a Deep Neural Network</a> -
<strong><em>Proceedings of the National Academy of
Sciences</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=11996680970579301810&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. David Bau’s review on network dissection for
discriminative and generative models.</p></li>
<li><p><a href="https://distill.pub/2020/circuits/zoom-in/">Zoom In: An
Introduction to Circuits</a> - <strong><em>Distill</em></strong>, 2020.
[<a
href="https://scholar.google.com/scholar?cluster=9053581372570691569&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A perspective on treating neural networks as
circuits.</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2020/hash/c74956ffb38ba48ed6ce977af6727275-Abstract.html">Compositional
Explanations of Neurons</a> - <strong><em>NeurIPS’20</em></strong>,
2020. [<a
href="https://scholar.google.com/scholar?cluster=15725346730266402738&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://github.com/jayelm/compexp">Project</a>]. A
concept-composition version of network dissection.</p></li>
<li><p><a
href="http://papers.NeurIPS.cc/paper/9095-this-looks-like-that-deep-learning-for-interpretable-image-recognition.pdf">This
Looks Like That: Deep Learning for Interpretable Image Recognition</a> -
<strong><em>NeurIPS’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=9461838581952136719&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.pnas.org/content/pnas/116/16/7723.full.pdf">Unsupervised
learning by competing hidden units</a> - <strong><em>Proceedings of the
National Academy of Sciences</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=1228003598355915526&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/2006.09994.pdf">Noise or Signal:
The Role of Backgrounds in Image Classification</a> -
<strong><em>ICLR’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=14729938011425134088&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://github.com/MadryLab/backgrounds_challenge">Code &amp;
Data</a>]. [<a
href="https://gradientscience.org/background/">Project</a>]. A
perspective on image background provides strong clue for foreground
classification.</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2018/hash/5fc34ed307aac159a30d81181c99847e-Abstract.html">Towards
Understanding Learning Representations: To What Extent Do Different
Neural Networks Learn the Same Representation</a> -
<strong><em>NeurIPS’18</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=401428033641216502&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Maching the learned pattern of neurons in different
neural networks.</p></li>
<li><p><a
href="https://kriegeskortelab.zuckermaninstitute.columbia.edu/sites/default/files/content/MehrerKietzmann_2020_NatureComms.pdf">Individual
differences among deep neural network models</a> - <strong><em>Nature
Communications</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=8259893575188417318&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="embodied-intelligence">Embodied Intelligence</h3>
<ul>
<li><p><a
href="https://plato.stanford.edu/entries/embodied-cognition/">Embodied
Cognition</a> - <strong><em>Plato Stanford</em></strong>. A
computational philosophy account on Embodied Cognition, which emphasizes
the significance of an agent’s physical body in cognitive
abilities.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/content-externalism/">Externalism
About the Mind</a> - <strong><em>Plato Stanford</em></strong>. A
computational philosophy account on mind externalism, a long-term debate
about the boundary of embodied intelligence.</p></li>
<li><p><a
href="https://www.researchgate.net/profile/David-Woods-19/publication/242545872_Cognitive_Engineering_Human_Problem_Solving_with_Tools/links/542becf70cf29bbc126ac097/Cognitive-Engineering-Human-Problem-Solving-with-Tools.pdf">Cognitive
engineering: Human problem solving with tools</a> - <strong><em>Human
Factors</em></strong>, 1988. [<a
href="https://scholar.google.com/scholar?cluster=14194840995416222723&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original idea of investigating huamn tool use in
problem solving.</p></li>
<li><p><a href="https://psycnet.apa.org/record/1993-97340-000">Tools,
language and cognition in human evolution</a> - <strong><em>Cambridge
University Press</em></strong>, 1993. [<a
href="https://scholar.google.com/scholar?cluster=6046350461147957220&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A classic perspective correlating human tool use with the
evolution of civilization.</p></li>
<li><p><a
href="https://icds.uoregon.edu/wp-content/uploads/2014/06/Clark-and-Chalmers-The-Extended-Mind.pdf">The
Extended Mind</a> - <strong><em>Analysis</em></strong>, 1998. [<a
href="https://scholar.google.com/scholar?cluster=9546561188261943866&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper on the debate of mind
externalism.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S1364661303003231">The
neural bases of complex tool use in humans</a> - <strong><em>Trends in
Cognitive Sciences</em></strong>, 2004. [<a
href="https://scholar.google.com/scholar?cluster=3612212926196611828&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A neuroscience account of human tool use.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0960982207017708">Spontaneous
Metatool Use by New Caledonian Crows</a> - <strong><em>Current
Biology</em></strong>, 2007. [<a
href="https://scholar.google.com/scholar?cluster=9263531730425342443&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A piece of evidence that intelligent animals can take
advantage of matatools to make tools for problem solving.</p></li>
<li><p><a
href="https://journals.sagepub.com/doi/abs/10.1177/0956797610371962">Rapid
Assimilation of External Objects Into the Body Schema</a> -
<strong><em>Psychological Science</em></strong>, 2010. [<a
href="https://scholar.google.com/scholar?cluster=854636910326733489&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.eva.mpg.de/documents/Cambridge/Tennie_Cultural_BehBrainSci_2012_1566208.pdf">The
cognitive bases of human tool use</a> - <strong><em>Behavioral and Brain
Sciences</em></strong>, 2012. [<a
href="https://scholar.google.com/scholar?cluster=4648150119820414671&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.frontiersin.org/articles/10.3389/fpsyg.2013.00214/full">The
embodied mind extended: using words as social tools</a> -
<strong><em>Frontiers in Psychology</em></strong>, 2013. [<a
href="https://scholar.google.com/scholar?cluster=14719988081062606352&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://royalsocietypublishing.org/doi/10.1098/rstb.2012.0408">Tool
use as adaptation</a> - <strong><em>Philosophical Transactions of the
Royal Society B: Biological Sciences</em></strong>, 2013. [<a
href="https://scholar.google.com/scholar?cluster=8060841461200774807&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0028393214000232">Intensive
tool-practice and skillfulness facilitate the extension of body
representations in humans</a> -
<strong><em>Neuropsychologia</em></strong>, 2014. [<a
href="https://scholar.google.com/scholar?cluster=10578024091098127929&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://psycnet.apa.org/doiLanding?doi=10.1037%2Frev0000027">Tool
use and affordance: Manipulation-based versus reasoning-based
approaches</a> - <strong><em>Psychological Review</em></strong>, 2016.
[<a
href="https://scholar.google.com/scholar?cluster=3284942486402374505&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A classic review on human tool use and
affordance.</p></li>
<li><p><a href="https://escholarship.org/uc/item/5gf0m7x3">Meta-strategy
learning in physical problem-solving: the effect of embodied
experience</a> - <strong><em>CogSci’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=9713842177532954702">All
Versions</a>].</p></li>
<li><p><a
href="https://yzhu.io/publication/tool2015cvpr/paper.pdf">Understanding
Tools: Task-Oriented Object Modeling, Learning and Recognition</a> -
<strong><em>CVPR’15</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=4609926671953500969&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://yzhu.io/publication/tool2015cvpr/">Project</a>]. The
original paper introducing affordance and physically-grounded tool use
into computer vision.</p></li>
<li><p><a
href="https://robotics.sciencemag.org/content/6/54/eabd7935.abstract">Robotic
hand augmentation drives changes in neural body representation</a> -
<strong><em>Science Robotics</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=1622125726197763917&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.jneurosci.org/content/jneuro/41/13/2980.full.pdf">Expert
Tool Users Show Increased Differentiation between Visual Representations
of Hands and Tools</a> - <strong><em>Journal of
Neuroscience</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=13454164767827515188&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/2106.05654.pdf">Visual scoping
operations for physical assembly</a> -
<strong><em>CogSci’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=7238090583833839&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.cc.gatech.edu/ai/robot-lab/online-publications/StoytchevICRA2005.pdf">Behavior-grounded
representation of tool affordances</a> -
<strong><em>ICRA’05</em></strong>, 2005. [<a
href="https://scholar.google.com/scholar?cluster=6115815663915603675&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://link.springer.com/chapter/10.1007/978-3-642-38812-5_1">A
Relational Approach to Tool-Use Learning in Robots</a> -
<strong><em>ILP’12</em></strong>, 2012. [<a
href="https://scholar.google.com/scholar?cluster=18374178227592386332&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007/s10514-017-9637-x">Relational
affordances for multiple-object manipulation</a> -
<strong><em>Autonomous Robots</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=6357646940615855682&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://m.roboticsproceedings.org/rss15/p01.pdf">Improvisation
through Physical Understanding: Using Novel Objects as Tools with Visual
Foresight</a> - <strong><em>RSS’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=4316276917607326251&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://escholarship.org/uc/item/5gf0m7x3">Meta-strategy
learning in physical problem-solving: the effect of embodied
experience</a> - <strong><em>CogSci’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=9713842177532954702">All
Versions</a>]. [<a href="">Preprint</a>]. This paper focuses on how
natural embodied experience affects what kinds of abstract physical
problem-solving strategies people use in a virtual task. The findings
suggest that differences in embodied experience drive the acquisition of
different meta-strategies for balancing acting with thinking, deciding
what kinds of actions to try, and deciding how persistent to be with a
current action plan.</p></li>
<li><p><a href="https://arxiv.org/abs/2002.06289">3D dynamic scene
graphs: Actionable spatial perception with places, objects, and
humans</a> - <strong><em>RSS’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=4428742298455436054&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A system for modeling 3D dynamic scene graphs on multiple
levels (metric-semantic mesh, objects and agents, places and structures,
rooms, and buildings).</p></li>
<li><p><a
href="https://www.nature.com/articles/s42256-025-01005-x">Embodied large
language models enable robots to complete complex tasks in unpredictable
environments</a> - <strong><em>Nature Machine
Intelligence</em></strong>, 2025. [<a
href="https://scholar.google.com/scholar?cluster=4507358987058849097">All
Versions</a>]. Completing complex tasks in unpredictable settings
challenges robotic systems, requiring a step change in machine
intelligence. Sensorimotor abilities are considered integral to human
intelligence. Thus, biologically inspired machine intelligence might
usefully combine artificial intelligence with robotic sensorimotor
capabilities. This work reports an embodied large-language-model-enabled
robot (ELLMER) framework, utilizing GPT-4 and a retrieval-augmented
generation infrastructure, to enable robots to complete long-horizon
tasks in unpredictable settings. The method extracts contextually
relevant examples from a knowledge base, producing action plans that
incorporate force and visual feedback and enabling adaptation to
changing conditions. The authors tested ELLMER on a robot tasked with
coffee making and plate decoration; these tasks consist of a sequence of
sub-tasks from drawer opening to pouring, each benefiting from distinct
feedback types and methods. The authors show that the ELLMER framework
allows the robot to complete the tasks. This demonstration marks
progress towards scalable, efficient and ‘intelligent robots’ able to
complete complex tasks in uncertain environments.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="evolutionary-intelligence">Evolutionary Intelligence</h3>
<ul>
<li><p><a
href="http://websites.umich.edu/~zhanglab/clubPaper/06_08_2012.pdf">Evolutionary
trade-offs, Pareto optimality, and the geometry of phenotype space</a> -
<strong><em>Science</em></strong>, 2012. [<a
href="https://scholar.google.com/scholar?cluster=16162252507845975080&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A classic paper correlating biological trade-offs with
the evolution of pareto optimality.</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007/BF01442131">Pareto
optimality in multiobjective problems</a> - <strong><em>Applied
Mathematics and Optimization</em></strong>, 1977. [<a
href="https://scholar.google.com/scholar?cluster=11305142600366783354&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper on the pareto optimality in
multiobjective problems.</p></li>
<li><p><a href="http://www.soft-computing.de/SMC0805.pdf">Pareto-Based
Multiobjective Machine Learning: An Overview and Case Studies</a> -
<strong><em>IEEE Transactions on Systems, Man, and
Cybernetics</em></strong>, 2008. [<a
href="https://scholar.google.com/scholar?cluster=11308312498510305429&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A comprehensive review on the application of pareto
optimality to multiobjective machine learning.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41586-019-1153-z">Phylogenetic
evidence for Sino-Tibetan origin in northern China in the Late
Neolithic</a> - <strong><em>Nature</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=13913123623752818925&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A Bayesian phylogenetic analysis on two competing
hypotheses of the origin of the Sino-Tibetan language family suggests
that the initial expansion of Sino-Tibetan languages occurred
approximately 4,000–6,000 years before present (BP; taken as AD 1950) in
the Yellow River basin of northern China, and that this expansion is
associated with the development of the Yangshao and/or Majiayao
Neolithic cultures.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41586-021-04108-8">Triangulation
supports agricultural spread of the Transeurasian languages</a> -
<strong><em>Nature</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=1183005894965630508&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://www.nature.com/articles/d41586-021-03037-w">Nature
News</a>]. A triangulation of linguistic, archaeological and genetic
data suggests that the Transeurasian language family originated in a
population of grain farmers in China around 9,000 years ago, and that
agriculture underpinned its spread.</p></li>
<li><p><a
href="https://www.science.org/doi/abs/10.1126/science.ade7981">From
language development to language evolution: A unified view of human
lexical creativity</a> - <strong><em>Science</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=15871163761816546924&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://brochhagen.github.io/content/ms/accepted-lexical-creativity.pdf">Preprint</a>].
This work supports a unified foundation for human lexical creativity
underlying both the fleeting products of individual ontogeny and the
evolutionary products of phylogeny across languages.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="methodologies-for-experiments">Methodologies for
Experiments</h3>
<h4 id="quantitative-analysis">Quantitative Analysis</h4>
<ul>
<li><p><a
href="http://www.jakebowers.org/ITVExperiments/angristimbensrubin96.pdf">Identification
of Causal Effects Using Instrumental Variables</a> - <strong><em>Journal
of the American Statistical Association</em></strong>, 1996. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=17166265099721941605">All
Versions</a>]. The original paper on Instrumental Variables for natural
sociology studies.</p></li>
<li><p><a
href="https://www.annualreviews.org/doi/abs/10.1146/annurev-psych-122414-033702">Experiments
with More Than One Random Factor: Designs, Analytic Models, and
Statistical Power</a> - <strong><em>Annual Review of
Psychology</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=6652444619934494760&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A comprehensive review of the quantitative analysis
techniques for behavioral studies.</p></li>
<li><p><a
href="https://mpra.ub.uni-muenchen.de/4823/1/MPRA_paper_4823.pdf">With
or Without U? The Appropriate Test for a U-Shaped Relationship</a> -
<strong><em>Oxford Bulletin of Economics and Statistics</em></strong>,
2010. [<a
href="https://scholar.google.com/scholar?cluster=1574723532506536904&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original method for testing U-shape relation from the
data, which is distinctive from the quadratic regression test.</p></li>
<li><p><a
href="https://journals.sagepub.com/doi/pdf/10.1177/2515245918805755">Two
lines: A valid alternative to the invalid testing of U-shaped
relationships with quadratic regressions</a> - <strong><em>Advances in
Methods and Practices in Psychological Science</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=12010185803500406162&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. An alternative method to test the statistical
significance of U-shaped relationships.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="scaling-up-behavioral-studies">Scaling Up Behavioral
Studies</h4>
<ul>
<li><p><a href="https://osf.io/wksv8">Scaling up experimental social,
behavioral, and economic science</a> - <strong><em>Open Science
Foundation Preprints</em></strong>. [<a
href="https://scholar.google.com/scholar?hl=en&amp;as_sdt=0%2C5&amp;q=Scaling+up+experimental+social%2C+behavioral%2C+and+economic+science&amp;btnG=">All
Versions</a>]. A white paper on scaling up social, behavioral, and
econimic experiments.</p></li>
<li><p><a
href="https://scholar.harvard.edu/files/henrich/files/henrich_heine_norenzayan_2010-2.pdf">The
weirdest people in the world?</a> - <strong><em>Brain and Behavioral
Sciences</em></strong>, 2010. [<a
href="https://scholar.google.com/scholar?cluster=3129419557801277936&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper on rethinking and tackling the sample
bias in behaivoral studies, where most subjects are drawn from Western,
Educated, Industrialized, Rich, and Democratic (WEIRD)
societies.</p></li>
<li><p><a
href="https://www.pnas.org/doi/10.1073/pnas.1915841117">Scaling up
psychology via Scientific Regret Minimization</a> -
<strong><em>Proceedings of the National Academy of
Sciences</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=8011895688226766944&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The statistical and ecological basis for scaling up
behavioral studies.</p></li>
<li><p><a
href="https://cpb-us-w2.wpmucdn.com/web.sas.upenn.edu/dist/a/511/files/2021/06/Bhatia-He-Science.pdf">Machine-generated
theories of human decision-making</a> -
<strong><em>Science</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=7065547001880027350&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://cocosci.princeton.edu/jpeterson/papers/peterson2021-science.pdf">Using
large-scale experiments and machine learning to discover theories of
human decision-making</a> - <strong><em>Science</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=7456250222852859810&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A piece of evidence for the merits brought by large-scale
behavioral studies in social science.</p></li>
<li><p><a
href="http://jakehofman.com/pdfs/integrating-prediction-and-explanation.pdf">Integrating
explanation and prediction in computational social science</a> -
<strong><em>Nature</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=288245575125750925&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://cocosci.princeton.edu/josh/papers/griffiths-largeimagedatabases-topics2016.pdf">Exploring
human cognition using large image databases</a> - <strong><em>Topics in
Cognitive Sciences</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=3629906005701226294&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://web.archive.org/web/20170809024454id_/http://www.kevinjing.com/visual_search_at_pinterest.pdf">Visual
Search at Pinterest</a> - <strong><em>KDD’15</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=2051024301293529405&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Large scale user study in the development of the
recommendations system by Pinterest.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="decision-making">Decision Making</h4>
<ul>
<li><a
href="https://link.springer.com/article/10.3758/s13428-022-01789-5">A
computational process-tracing method for measuring people’s planning
strategies and how they change over time</a> - <strong><em>Behavior
Research Methods</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=10405935000926098041">All
Versions</a>]. Model-based strategy identification.</li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="question-answering">Question Answering</h4>
<ul>
<li><p><a
href="https://cogsci.mindmodeling.org/2016/papers/0122/paper0122.pdf">Searching
large hypothesis spaces by asking questions</a> -
<strong><em>CogSci’16</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=3398849603439166012&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A behavioral study for the 20 questions game.</p></li>
<li><p><a
href="https://gureckislab.org/papers/RotheLakeGureckis-2016cogsci.pdf">Asking
and evaluating natural language questions</a> -
<strong><em>CogSci’16</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=34641833161282231&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A behavioral study for the battleship game.</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007/s42113-018-0005-5">Do
People Ask Good Questions?</a> - <strong><em>Computational Brain &amp;
Behavior</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=14595996621617337270&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://nyuccl.org/papers/Rothe-Lake-Gureckis-2019-Cogsci.pdf">Asking
goal-oriented questions and learning from answers</a> -
<strong><em>CogSci’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=14185546187726917682&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="human-machine-comparison">Human-Machine Comparison</h4>
<ul>
<li><p><a
href="https://psycnet.apa.org/record/1973-00249-001">Elimination by
aspects: A theory of choice</a> - <strong><em>Psychological
Review</em></strong>, 1972. [<a
href="https://scholar.google.com/scholar?cluster=1633792484482810297&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Herbert Simon’s early experiments on computer aided
behavioral studies.</p></li>
<li><p><a
href="https://stacks.stanford.edu/file/druid:qv796fc9687/qv796fc9687.pdf">Problem
Solving and Rule Induction: A Unified View</a> - <strong><em>Knowledge
and cognition</em></strong>, 1974. [<a
href="https://scholar.google.com/scholar?cluster=12943734683291006234&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.pnas.org/content/112/37/11708.short">Evidence
integration in model-based tree search</a> - <strong><em>Proceedings of
the National Academy of Sciences</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=11085043350027609187&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://link.springer.com/content/pdf/10.1007/s42113-019-00053-y.pdf">People
Infer Recursive Visual Concepts from Just a Few Examples</a> -
<strong><em>Computational Brain &amp; Behavior</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=3871396883970734141&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://escholarship.org/content/qt3xf2n3vc/qt3xf2n3vc.pdf">One-shot
learning of generative speech concepts</a> -
<strong><em>CogSci’14</em></strong>, 2014. [<a
href="https://scholar.google.com/scholar?cluster=15482292457660075957&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/1901.04587">Human few-shot
learning of compositional instructions</a> -
<strong><em>CogSci’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=12841163907815018136&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/2103.05823.pdf">Fast and flexible:
Human program induction in abstract reasoning tasks</a> -
<strong><em>CogSci’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=5294483826040237516&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://proceedings.mlr.press/v80/dubey18a.html">Investigating
Human Priors for Playing Video Games</a> -
<strong><em>ICML’18</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=2202192690517876762&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S2352154619300622">Tasks
for aligning human and machine planning</a> - <strong><em>Current
Opinion in Behavioral Sciences</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=8308872468787875598&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://perception.jhu.edu/files/PDFs/19_Adversarial_Deciphering/ZhouFirestone-AdversarialDeciphering.pdf">Humans
can decipher adversarial images</a> - <strong><em>Nature
Communications</em></strong>. 2019. [<a
href="https://scholar.google.com/scholar?cluster=4423950118844131054&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.nature.com/articles/s41593-022-01026-4.pdf">Shared
computational principles for language processing in humans and deep
language models</a> - <strong><em>Nature Neuroscience</em></strong>,
2022. [<a
href="https://scholar.google.com/scholar?cluster=16078004657063602593&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="association-test">Association Test</h4>
<ul>
<li><p><a
href="https://en.wikipedia.org/wiki/Implicit-association_test">Implicit
Association Test</a> - <strong><em>Wikipedia</em></strong>. Wikipedia on
the Implicit Association Test, a controversial assessment intended to
detect subconscious associations between mental representations of
objects (concepts) in memory.</p></li>
<li><p><a
href="http://faculty.fortlewis.edu/burke_b/Senior/BLINK%20replication/IAT.pdf">Measuring
Individual Differences in Implicit Cognition: The Implicit Association
Test</a> - <strong><em>Journal of Personality and Social
Psychology</em></strong>, 1998. [<a
href="https://scholar.google.com/scholar?cluster=302378224541015580&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper introducing the Implicit Association
Test.</p></li>
<li><p><a
href="http://faculty.washington.edu/agg/pdf/Gwald_Nosek_ZEITSCHR_2001.OCR.pdf">Health
of the Implicit Association Test at age 3</a> - <strong><em>Zeitschrift
für Experimentelle Psychologie</em></strong>, 2001. [<a
href="https://scholar.google.com/scholar?cluster=10868478693422595588&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The 3rd year review for the IAT.</p></li>
<li><p><a
href="https://faculty.washington.edu/agg/pdf/Nosek%20&amp;%20al.IATatage7.2007.pdf">The
Implicit Association Test at Age 7: A Methodological and Conceptual
Review</a> - <strong><em>Social psychology and the unconscious: The
automaticity of higher mental processes (pp. 265–292), Psychology
Press</em></strong>, 2007. [<a
href="https://scholar.google.com/scholar?cluster=16189750920013376566&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The 7th year review for the IAT.</p></li>
<li><p><a
href="http://faculty.washington.edu/agg/IATmaterials/PDFs/Hofmann%20&amp;%20al%20(PSPB,2005).pdf">A
Meta-Analysis on the Correlation Between the Implicit Association Test
and Explicit Self-Report Measures</a> - <strong><em>Personality and
Social Psychology Bulletin</em></strong>, 2005. [<a
href="https://scholar.google.com/scholar?cluster=4888328728717829047&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="virtual-reality">Virtual Reality</h4>
<ul>
<li><p><a href="https://www.nature.com/articles/nn948">Virtual reality
in behavioral neuroscience and beyond</a> - <strong><em>Nature
Neuroscience</em></strong>, 2002. [<a
href="https://scholar.google.com/scholar?cluster=12168354203281280346&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A classic review on the early applications of Virtual
Reality to behavioral studies.</p></li>
<li><p><a
href="https://stanfordvr.com/mm/2009/fox-jmp-vr-survival.pdf">Virtual
reality: A survival guide for the social scientist</a> -
<strong><em>Journal of Media Psychology</em></strong>, 2009. [<a
href="https://scholar.google.com/scholar?cluster=17318470193315023264&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://psycnet.apa.org/record/2022-60836-006">The
psychology of virtual reality</a> - <strong><em>The psychology of
technology: Social science research in the age of Big Data
(pp. 155–193), American Psychological Association</em></strong>, 2022.
[<a
href="https://scholar.google.com/scholar?cluster=11535480055596209683&amp;hl=en&amp;as_sdt=0,5&amp;as_ylo=2021">All
Versions</a>]. Jeremy Bailenson’s review on the applications of Virtual
Reality to behavioral studies.</p></li>
<li><p><a
href="https://stanfordvr.com/mm/2015/cummings-mp-how-immersive.pdf">How
Immersive Is Enough? A Meta-Analysis of the Effect of Immersive
Technology on User Presence</a> - <strong><em>Media
Psychology</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=9218122072360464558&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A meta-analysis on the extent to which technologies need
to be immersive in order to generate a sense of presence.</p></li>
<li><p><a href="https://ieeexplore.ieee.org/document/10108427">Towards
an Understanding of Distributed Asymmetric Collaborative Visualization
on Problem-solving</a> - <strong><em>VR’23</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=11228377215337222005&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://dl.acm.org/doi/abs/10.1145/3139131.3139152">Agent:
automatic generation of experimental protocol runtime</a> -
<strong><em>VRST’17</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=3511549412244980073">All
Versions</a>]. This paper proposes the use of Domain-Specific Languages
(DSLs) to ease the description and generation of VR experiments, thus
letting experiment designers focus on their core tasks: designing,
conducting, and reporting experiments.</p></li>
<li><p><a
href="https://dl.acm.org/doi/abs/10.1145/3654777.3676358">What’s the
Game, then? Opportunities and Challenges for Runtime Behavior
Generation</a> - <strong><em>UIST’24</em></strong>, 2024. [<a
href="">All Versions</a>]. Procedural content generation (PCG), the
process of algorithmically creating game components instead of manually,
has been a common tool of game development for decades. Recent advances
in large language models (LLMs) enable the generation of game behaviors
based on player input at runtime. Such code generation brings with it
the possibility of entirely new gameplay interactions that may be
difficult to integrate with typical game development workflows. This
work explores these implications through GROMIT, a novel LLM-based
runtime behavior generation system for Unity. When triggered by a player
action, GROMIT generates a relevant behavior which is compiled without
developer intervention and incorporated into the game.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="meta-level-considerations">Meta-Level Considerations</h3>
<h4 id="meta-learning">Meta Learning</h4>
<ul>
<li><p><a href="https://arxiv.org/pdf/2201.03916.pdf">Automated
Reinforcement Learning (AutoRL): A Survey and Open Problems</a> - 2022.
[<a
href="https://scholar.google.com/scholar?cluster=9025378857688824887&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A comprehensive review on AutoRL.</p></li>
<li><p><a
href="https://proceedings.mlr.press/v70/finn17a/finn17a.pdf">Model-Agnostic
Meta-Learning for Fast Adaptation of Deep Networks</a> -
<strong><em>ICML’17</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=17278604844873996878">All
Versions</a>]. [<a
href="https://bair.berkeley.edu/blog/2017/07/18/learning-to-learn/">Post</a>].
Chelsea Finn’s original paper on Model-Agnostic Meta-Learning
(MAML).</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2018/hash/e1021d43911ca2c1845910d84f40aeae-Abstract.html">Bayesian
Model-Agnostic Meta-Learning</a> - <strong><em>NeurIPS’18</em></strong>,
2018. [<a
href="https://scholar.google.com/scholar?cluster=7370333111335795917&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A Bayesian account on MAML.</p></li>
<li><p><a
href="https://openreview.net/forum?id=SJeD3CEFPH">Meta-Q-Learning</a> -
<strong><em>ICLR’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=2865388954464396222&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The milestone paper on context Meta-RL.</p></li>
<li><p><a
href="http://proceedings.mlr.press/v97/rakelly19a.html">Efficient
Off-Policy Meta-Reinforcement Learning via Probabilistic Context
Variables</a> - <strong><em>ICML’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=15379570585451726919&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://openreview.net/forum?id=TQt98Ya7UMP">Balancing
Constraints and Rewards with Meta-Gradient D4PG</a> -
<strong><em>ICLR’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=2805226315118298313&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://openreview.net/forum?id=Bk8BvDqex">Metacontrol
for Adaptive Imagination-Based Optimization</a> -
<strong><em>ICLR’17</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=16728474512617398730&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2021/hash/1e4d36177d71bbb3558e43af9577d70e-Abstract.html">On
Effective Scheduling of Model-based Reinforcement Learning</a> -
<strong><em>NeurIPS’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=11128521607771619105&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="marrs-levels-of-analysis">Marr’s Levels of Analysis</h4>
<ul>
<li><p><a href="https://usa1lib.org/book/1223444/8e5ca8">Vision: A
Computational Investigation into the Human Representation and Processing
of Visual Information</a> - <strong><em>MIT Press</em></strong>, 1982.
[<a
href="https://scholar.google.com/scholar?cluster=14386368570811483142&amp;hl=en&amp;as_sdt=0,44">All
Versions</a>]. David Marr’s original book on the levels of
analysis.</p></li>
<li><p><a
href="https://dspace.mit.edu/bitstream/handle/1721.1/5782/AIM-357.pdf?sequence=2">From
understanding computation to understanding neural circuitry</a> -
<strong><em>Neuroscience Research Program Bulletin</em></strong>, 1979.
[<a
href="https://scholar.google.com/scholar?start=0&amp;hl=en&amp;as_sdt=0,5&amp;cluster=11150567121969913334">All
Versions</a>].</p></li>
<li><p><a
href="https://cocosci.princeton.edu/tom/papers/LabPublications/BridgingLevelsAnalysis.pdf">Bridging
Levels of Analysis for Probabilistic Models of Cognition</a> -
<strong><em>Current Directions in Psychological Science</em></strong>,
2012. [<a
href="https://scholar.google.com/scholar?cluster=5063382112136991296&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A Marr’s paradigm account on probabilistic
models.</p></li>
<li><p><a
href="https://people.csail.mit.edu/pkrafft/papers/krafft-griffiths-levels-css.pdf">Levels
of Analysis in Computational Social Science</a> -
<strong><em>CogSci’18</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=10178929388985626844&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A Marr’s paradigm account on computational social
science.</p></li>
<li><p><a href="https://baicsworkshop.github.io/pdf/BAICS_6.pdf">Levels
of Analysis for Machine Learning</a> - <strong><em>ICLR’20 Bridging AI
and Cognitive Science Workshop</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=13819038971626384115&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A Marr’s paradigm account on machine learning.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="gestalt">Gestalt</h4>
<ul>
<li><p><a href="https://psycnet.apa.org/record/2007-10344-001">Gestalt
theory</a> - <strong><em>A source book of Gestalt
psychology</em></strong>, 1938. [<a
href="https://scholar.google.com/scholar?cluster=18133275659218646817&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original book on Gestalt psychology.</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007/BF00422382">Gestalt
Psychology</a> - <strong><em>Psychologische Forschung</em></strong>,
1967. [<a
href="https://scholar.google.com/scholar?cluster=16023098380090751616&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Wolfgang Köhler’s review on Gestalt psychology.</p></li>
<li><p><a
href="https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1467-9450.1984.tb01001.x">Restructuring
revisited I. Summary and critique of the Gestalt theory of problem
solving</a> - <strong><em>Scandinavian Journal of
Psychology</em></strong>, 1984. [<a
href="https://scholar.google.com/scholar?cluster=1540079499182933565&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1467-9450.1984.tb01005.x">Restructuring
revisited II. An information processing theory of restructuring and
insight</a> - <strong><em>Scandinavian Journal of
Psychology</em></strong>, 1984. [<a
href="https://scholar.google.com/scholar?cluster=1821980539002417470&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://psycnet.apa.org/record/1993-36184-001">Thoughts
beyond words: When language overshadows insight</a> -
<strong><em>Journal of Experimental Psychology</em></strong>, 1993. [<a
href="https://scholar.google.com/scholar?cluster=13773440938721955384&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://hk1lib.org/book/1244721/20ddc5">Deep Learning:
How the Mind Overrides Experience</a> - <strong><em>Cambridge University
Press</em></strong>, 2011. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=231021877034210140">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="the-aha-moment">The Aha! Moment</h4>
<ul>
<li><p><a href="https://en.wikipedia.org/wiki/Eureka_effect">Eureka
Effect</a> - <strong><em>Wikipedia</em></strong>. Wikipedia on Eureka
effect (a.k.a. Aha! moment, insight, and epiphany), the common human
experience of suddenly understanding a previously incomprehensible
problem or concept.</p></li>
<li><p><a href="https://en.wikipedia.org/wiki/Insight">Insight</a> -
<strong><em>Wikipedia</em></strong>. Wikipedia on insight.</p></li>
<li><p><a
href="https://en.wikipedia.org/wiki/Epiphany_(feeling)">Epiphany</a> -
<strong><em>Wikipedia</em></strong>. Wikipedia on epiphany, the
“feeling” when the Aha! moment comes.</p></li>
<li><p><a href="https://escholarship.org/uc/item/54x8v354">A
computational model of scientific insight</a> - <strong><em>The nature
of creativity: Contemporary psychological perspectives</em></strong>,
1988. [<a
href="https://scholar.google.com/scholar?cluster=13633357571064621019&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A computational account on insights for scientific
discovery.</p></li>
<li><p><a
href="https://www.researchgate.net/profile/Thomas-Ormerod/publication/8909475_What_Makes_an_Insight_Problem_The_Roles_of_Heuristics_Goal_Conception_and_Solution_Recoding_in_Knowledge-Lean_Problems/links/00b7d5159f3c057eb5000000/What-Makes-an-Insight-Problem-The-Roles-of-Heuristics-Goal-Conception-and-Solution-Recoding-in-Knowledge-Lean-Problems.pdf">What
Makes an Insight Problem? The Roles of Heuristics, Goal Conception, and
Solution Recoding in Knowledge-Lean Problems</a> - <strong><em>Journal
of Experimental Psychology</em></strong>, 2004. [<a
href="https://scholar.google.com/scholar?cluster=17529631069707671285&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://psycnet.apa.org/record/2003-10949-002">APA</a>].</p></li>
<li><p><a
href="https://www.hf.uni-koeln.de/data/fgpsych/File/Haider/Knoblich_etal_1999.pdf">Constraint
relaxation and chunk decomposition in insight problem solving</a> -
<strong><em>Journal of Experimental Psychology</em></strong>, 1999. [<a
href="https://scholar.google.com/scholar?cluster=8057214169831054227&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://psycnet.apa.org/record/1999-01477-011">APA</a>].</p></li>
<li><p><a
href="https://citeseerx.ist.psu.edu/document?repid=rep1&amp;type=pdf&amp;doi=818fec7c896ea3716eeb637da095293e9e6d1806">Dynamics
and constraints in insight problem solving</a> - <strong><em>Journal of
Experimental Psychology</em></strong>, 2002. [<a
href="https://scholar.google.com/scholar?cluster=12067671710370549516&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://psycnet.apa.org/record/2002-01361-014">APA</a>].</p></li>
<li><p><a
href="https://bpb-us-e1.wpmucdn.com/sites.northwestern.edu/dist/a/699/files/2015/11/Salvi_etal_Insight-is-right_TR2016-2n3ns9l.pdf">Insight
solutions are correct more often than analytic solutions</a> -
<strong><em>Thinking &amp; Reasoning</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=883561570778414219&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://docs.lib.purdue.edu/cgi/viewcontent.cgi?article=1094&amp;context=jps">Human
Performance on Insight Problem Solving: A Review</a> - <strong><em>The
Journal of Problem Solving</em></strong>, 2011. [<a
href="https://scholar.google.com/scholar?cluster=15913242870565808883&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.frontiersin.org/articles/10.3389/fpsyg.2016.01424/full">Insight
Is Not in the Problem: Investigating Insight in Problem Solving across
Task Types</a> - <strong><em>Frontiers in Psychology</em></strong>,
2016. [<a
href="https://scholar.google.com/scholar?cluster=4564128114316001308&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.researchgate.net/profile/Trina-Kershaw/publication/8909474_Multiple_Causes_of_Difficulty_in_Insight_The_Case_of_the_Nine-Dot_Problem/links/55dca27e08aeb38e8a8d23b6/Multiple-Causes-of-Difficulty-in-Insight-The-Case-of-the-Nine-Dot-Problem.pdf">Multiple
Causes of Difficulty in Insight: The Case of the Nine-Dot Problem</a> -
<strong><em>Journal of Experimental Psychology</em></strong>, 2004. [<a
href="https://scholar.google.com/scholar?cluster=15600199808825346018&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://psycnet.apa.org/record/2003-10949-001">APA</a>].</p></li>
<li><p><a
href="https://www.researchgate.net/profile/Gary-Jones-14/publication/23152585_Investigating_the_Effect_of_Mental_Set_on_Insight_Problem_Solving/links/0fcfd50abb767b1102000000/Investigating-the-Effect-of-Mental-Set-on-Insight-Problem-Solving.pdf">Investigating
the effect of Mental Set on Insight Problem Solving</a> -
<strong><em>Experimental Psychology</em></strong>, 2008. [<a
href="https://scholar.google.com/scholar?cluster=11054712671934144981&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="rationality">Rationality</h4>
<ul>
<li><p><a
href="https://plato.stanford.edu/entries/bounded-rationality/">Bounded
Rationality</a> - <strong><em>Plato Stanford</em></strong>. A
computational philosophy account on Bounded Rationality, an elementary
hypothesis of human intelligence in psychology and ecology.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/rationality-instrumental/">Instrumental
Rationality</a> - <strong><em>Plato Stanford</em></strong>. A
computational philosophy account on Instrumental Rationality, a dabate
on whether an agent’s decision is made intentionally or out of rational
coherence.</p></li>
<li><p><a
href="https://www.taylorfrancis.com/books/mono/10.4324/9781315083223/study-thinking-jerome-bruner-jacqueline-goodnow-george-austin">A
Study of Thinking</a> - <strong><em>Routledge</em></strong>, 1956. [<a
href="https://scholar.google.com/scholar?cluster=17466297915128086930">All
Versions</a>]. This book is a pioneering account of how human beings
achieve a measure of rationality in spite of the constraints imposed by
time and ignorance.</p></li>
<li><p><a
href="http://act-r.psy.cmu.edu/wordpress/wp-content/uploads/2012/12/89AdaptiveNature.pdf">The
Adaptive Nature of Human Categorization Behavior</a> -
<strong><em>Psychological Review</em></strong>, 1991. [<a
href="https://scholar.google.com/scholar?cluster=7349048316173616836&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper that relates cognitive resource
limitation with Bayesian rational analysis, in the case of
categorization behavior.</p></li>
<li><p><a
href="https://www.cell.com/trends/cognitive-sciences/fulltext/S1364-6613(03)00028-7?large_figure=true&amp;mobileUi=0">Task
switching</a> - <strong><em>Trends in Cognitive Sciences</em></strong>,
2003. [<a
href="https://scholar.google.com/scholar?cluster=676255515965300942&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="http://psychfiles.net/experimental/Monsell_2003.pdf">Preprint</a>].
The original paper on ``switch cost’‘, where subjects’ responses are
substantially slower and, usually, more error-prone immediately after a
task switch.</p></li>
<li><p><a
href="https://onlinelibrary.wiley.com/doi/full/10.1111/tops.12086">Computational
Rationality: Linking Mechanism and Behavior Through Bounded Utility
Maximization</a> - <strong><em>Topics in Cognitive
Science</em></strong>, 2014. [<a
href="https://scholar.google.com/scholar?cluster=15813211310327194798&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Introducing the computational rationality framework for
including information-processing bounds in rational analyses, which
emphasizes the incorporation of computational mechanism into the
definition of rational action.</p></li>
<li><p><a
href="https://gershmanlab.com/pubs/GershmanHorvitzTenenbaum15.pdf">Computational
rationality: A converging paradigm for intelligence in brains, minds,
and machines</a> - <strong><em>Science</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=7744057022238735461&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A comprehensive review on the rationality of Bayesian
computational models.</p></li>
<li><p><a
href="https://cocosci.princeton.edu/papers/lieder_resource.pdf">Resource-rational
analysis: Understanding human cognition as the optimal use of limited
computational resources</a> - <strong><em>Behavioral and Brain
Sciences</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=1642626865293965288&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A resource-rational account on interpreting human
intelligence.</p></li>
<li><p><a
href="https://onlinelibrary.wiley.com/doi/full/10.1111/tops.12142">Rational
Use of Cognitive Resources: Levels of Analysis Between the Computational
and the Algorithmic</a> - <strong><em>Topics in Cognitive
Science</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=16305499937147933368&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. An earlier version of the paper above.</p></li>
<li><p><a
href="https://www.cell.com/trends/cognitive-sciences/fulltext/S1364-6613(20)30215-1">Understanding
Human Intelligence through Human Limitations</a> - <strong><em>Trends in
Cognitive Sciences</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=6469796133334580403">All
Versions</a>]. [<a
href="https://cocosci.princeton.edu/papers/griffiths_understanding.pdf">Preprint</a>].
Recent progress in artificial intelligence provides the opportunity to
ask the question of what is unique about human intelligence, but with a
new comparison class. The author argues that we can understand human
intelligence, and the ways in which it may differ from artificial
intelligence, by considering the characteristics of the kind of
computational problems that human minds have to solve. The author claims
that these problems acquire their structure from three fundamental
limitations that apply to human beings: limited time, limited
computation, and limited communication. From these limitations we can
derive many of the properties we associate with human intelligence, such
as rapid learning, the ability to break down problems into parts, and
the capacity for cumulative cultural evolution.</p></li>
<li><p><a
href="https://eccl.mit.edu/s/Pelz_Foundations-of-intuitive-power-analyses-in-children-and-adults.pdf">Foundations
of intuitive power analyses in children and adults</a> -
<strong><em>Nature Human Behavior</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=4370839893505978405&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Evidences support that people have some of the
foundations for ‘intuitive power analyses’, which help people use
intuitive statistical reasoning and metacognitive strategies to estimate
how much information they might need to solve different discrimination
problems.</p></li>
<li><p><a
href="https://cocosci.princeton.edu/papers/ho2022cognitive.pdf">Cognitive
Science as a Source of Forward and Inverse Models of Human Decisions for
Robotics and Control</a> - <strong><em>Annual Review of Control,
Robotics, and Autonomous Systems</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;cluster=14055765901243029337">All
Versions</a>]. The review focuses on how cognitive science can provide
forward models of human decision-making and inverse models of how humans
think about others’ decision-making. The authors highlight relevant
recent developments, including approaches that synthesize black box and
theory-driven modeling, accounts that recast heuristics and biases as
forms of bounded optimality, and models that characterize human theory
of mind and communication in decision-theoretic terms.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="cognitive-architecture">Cognitive Architecture</h4>
<ul>
<li><p><a
href="https://plato.stanford.edu/entries/epistemology/">Epistemology</a>
- <strong><em>Plato Stanford</em></strong>.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S1364661321001285">The
secret life of predictive brains: what’s spontaneous activity for?</a> -
<strong><em>Trends in Cognitive Sciences</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=719229834892860829&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A neuroscience account on brain as a generative
model.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/abs/pii/0004370287900506">SOAR:
An architecture for general intelligence</a> - <strong><em>Artificial
Intelligence</em></strong>, 1987. [<a
href="https://scholar.google.com/scholar?cluster=10873259207109132615&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://act-r.psy.cmu.edu/wordpress/wp-content/uploads/2013/09/Anderson91.pdf">Is
human cognition adaptive?</a> - <strong><em>Behavioral and Brain
Sciences</em></strong>, 1991. [<a
href="https://scholar.google.com/scholar?cluster=3639936076538071052&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper introducing the adaptation perspective
of human intelligence, the theoretical basis of the ACT cognitive
architecture.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0004370205001530">Metacognition
in computation: A selected research review</a> - <strong><em>Artificial
Intelligence</em></strong>, 2005. [<a
href="https://scholar.google.com/scholar?cluster=4240334051245008914&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0010027718301604">Basic
functional trade-offs in cognition: An integrative framework</a> -
<strong><em>Cognition</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=11475742130443069967&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://doi.org/10.1126/SCIENCE.AAN8871">What is
consciousness, and could machines have it?</a> -
<strong><em>Science</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=6932714857132107942&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A perspective on the two levels of consciousness in
machine intelligence.</p></li>
<li><p><a
href="https://www.worldscientific.com/doi/abs/10.1142/S2705078521500028">A
Theoretical Computer Science Perspective on Consciousness</a> -
<strong><em>Journal of Artificial Intelligence and
Consciousness</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=16430561748075101972&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="science-logology">Science Logology</h3>
<h4 id="philosophy-of-science">Philosophy of Science</h4>
<ul>
<li><p><a
href="https://www-inst.eecs.berkeley.edu/~cs298-7/fa20/readings/kuhn.pdf">The
structure of scientific revolutions</a> - <strong><em>University of
Chicago Press: Chicago</em></strong>, 1970. [<a
href="https://scholar.google.com/scholar?cluster=8909475038284903063&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Thomas Kuhn’s original book on the emergence and the
shift of scientific paradigms.</p></li>
<li><p><a href="https://jamacoartney.net/Abend%20(2008).pdf">The Meaning
of “Theory”</a> - <strong><em>Sociological Theory</em></strong>, 2008.
[<a
href="https://scholar.google.com/scholar?cluster=4876642889050563131&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A philosophical account on the definition of “theory” in
social science (also can be generalized to natural science).</p></li>
<li><p><a
href="https://journals.sagepub.com/doi/pdf/10.4256/mio.2013.015">The
blind men and the elephant: A metaphor to illuminate the role of
researchers and reviewers in social science</a> -
<strong><em>Methodological Innovations Online</em></strong>, 2013. [<a
href="https://scholar.google.com/scholar?cluster=1654629562068006152&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://dl.acm.org/doi/abs/10.1145/3576896">A
Computational Inflection for Scientific Discovery</a> -
<strong><em>Communications of the ACM</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=1756108647531090189&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="science-of-science">Science of Science</h4>
<ul>
<li><p><a
href="https://en.wikipedia.org/wiki/Metascience">Metascience</a> -
<strong><em>Wikipedia</em></strong>.</p></li>
<li><p><a
href="http://ctbergstrom.com/publications/pdfs/2018Science.pdf">Science
of Science</a> - <strong><em>Science</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=6471468823556848055&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A comprehensive large-scale review on the science of
science.</p></li>
<li><p><a
href="https://www.pnas.org/doi/abs/10.1073/pnas.0307752101">Finding
Scientific Topics</a> - <strong><em>Proceedings of the National Academy
of Sciences</em></strong>, 2004. [<a
href="https://scholar.google.com/scholar?cluster=17382767110929995134&amp;hl=zh-CN&amp;as_sdt=0,5">All
Versions</a>]. Thomas L. Griffiths’s analysis of scientific topics using
Bayesian model.</p></li>
<li><p><a
href="https://www.pnas.org/doi/10.1073/pnas.1618569114">Meta-assessment
of Bias in Science</a> - <strong><em>Proceedings of the National Academy
of Sciences</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=14575889060982308028&amp;hl=zh-CN&amp;as_sdt=0,5">All
Verisions</a>]. An analysis of bias patterns and risk factors in
science.</p></li>
<li><p><a href="https://www.pnas.org/doi/10.1073/pnas.2021636118">Slowed
Canonical Progress in Large Fields of Science</a> -
<strong><em>Proceedings of the National Academy of
Sciences</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=7541922918797308487&amp;hl=zh-CN&amp;as_sdt=0,5">All
Verisions</a>]. An analysis of why too many papers published each year
in a field can lead to stagnation rather than advance.</p></li>
<li><p><a href="https://dl.acm.org/doi/10.1145/2858036.2858283">HCI
Research as Problem-Solving</a> - <strong><em>ACM
SIGCHI’16</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=3206201064123443333&amp;as_sdt=0,5">All
Versions</a>]. This essay contributes a meta-scientific account of
human-computer interaction (HCI) research as problem-solving. We build
on the philosophy of Larry Laudan, who develops problem and solution as
the foundational concepts of science. We argue that most HCI research is
about three main types of problem: empirical, conceptual, and
constructive.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="literature-mining">Literature Mining</h4>
<ul>
<li><p><a
href="https://www.nature.com/articles/s41467-024-45563-x">Structured
information extraction from scientific text with large language
models</a> - <strong><em>Nature Communications</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=13694008040033857249">All
Versions</a>]. This paper presents a simple approach to joint named
entity recognition and relation extraction and demonstrate how
pretrained large language models can be fine-tuned to extract useful
records of complex scientific knowledge. The authors test three
representative tasks in materials chemistry: linking dopants and host
materials, cataloging metal-organic frameworks, and general
composition/phase/morphology/application information
extraction.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41467-020-17266-6">Automated
extraction of chemical synthesis actions from experimental
procedures</a> - <strong><em>Nature Communications</em></strong>, 2020.
[<a
href="https://scholar.google.com/scholar?cluster=1626689948540815082">All
Versions</a>]. This paper presents a method to convert unstructured
experimental procedures written in English to structured synthetic steps
(action sequences) reflecting all the operations needed to successfully
conduct the corresponding chemical reactions.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41467-021-22951-1">Inferring
experimental procedures from text-based representations of chemical
reactions</a> - <strong><em>Nature Communications</em></strong>, 2021.
[<a
href="https://scholar.google.com/scholar?cluster=15772647675166217556">All
Versions</a>]. This paper presents data-driven models for predicting the
entire sequence of synthesis steps starting from a textual
representation of a chemical equation, for application in batch organic
chemistry.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41467-023-43836-5">Language
models and protocol standardization guidelines for accelerating
synthesis planning in heterogeneous catalysis</a> - <strong><em>Nature
Communications</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=8186755371438552520">All
Versions</a>]. This paper introduces a transformer model for automated
synthesis protocol analysis in catalyst discovery, exemplified using
single-atom heterogeneous catalysts (SACs), a rapidly expanding catalyst
family. The model adeptly converts SAC protocols into action sequences,
and this output is used to facilitate statistical inference of their
synthesis trends and applications, potentially expediting literature
review and analysis.</p></li>
<li><p><a href="https://www.nature.com/articles/s41598-025-02643-2">An
intelligent guided troubleshooting method for aircraft based on
HybirdRAG</a> - <strong><em>Scientific Reports</em></strong>, 2025. [<a
href="https://scholar.google.com/scholar?cluster=4924119792997395046">All
Versions</a>]. To enhance aircraft fault diagnosis efficiency, this
paper proposes HybridRAG, an intelligent-guided troubleshooting
framework that integrates knowledge graphs and large language models
(LLMs). Unlike conventional retrieval-augmented generation (RAG) methods
that rely on single-modal retrieval, HybridRAG adopts a
multi-dimensional retrieval strategy, combining graph-based reasoning
with both vector-based and BM25-based text retrieval techniques. This
hybrid approach ensures comprehensive extraction of relevant information
from both unstructured text and structured fault graphs, enhancing
diagnostic precision, relevance, and robustness. Experimental results
demonstrate that HybridRAG achieves an F1 score improvement of at least
4% and reduces hallucination rates by over 7% compared to mainstream RAG
baselines. These advancements, combined with its unique integration of
multi-modal retrieval, position HybridRAG as a novel framework for
addressing complex aircraft maintenance challenges. Additionally, the
paper presents an agent-based intelligent troubleshooting assistant that
supports more interactive, adaptive, and flexible diagnostic Q&amp;A,
providing maintenance personnel with a significant advanced intelligent,
context-aware diagnostic tool.</p></li>
<li><p><a href="https://galactica.org/static/paper.pdf">Galactica: A
Large Language Model for Science</a> - <strong><em>Meta
AI</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=15782429788006956926&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A large language model trained on large-scale scientific
corpus.</p></li>
<li><p><a href="https://arxiv.org/abs/2205.03512">CORWA: A
Citation-Oriented Related Work Annotation Dataset</a> -
<strong><em>NAACL’22</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=14605899782190710454&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://aclanthology.org/2021.acl-demo.14/">ESRA:
Explainable Scientific Research Assistant</a> - <strong><em>ACL’21 Demo
Track</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=4387915912582172679&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A tool for constructing and visualizing the knowledge
graph of a query keyword in literature retrieving.</p></li>
<li><p><a
href="https://matthewberger.github.io/papers/cite2vec.pdf">cite2vec:
Citation-Driven Document Exploration via Word Embeddings</a> -
<strong><em>IEEE Transactions on Visualization and Computer
Graphics</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=6949650208780085923&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://cic.tju.edu.cn/faculty/zhangjiawan/Jiawan_Zhang_files/paper/zeyuli2020.pdf">Galex:
Exploring the evolution and intersection of disciplines</a> -
<strong><em>IEEE Transactions on Visualization and Computer
Graphics</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=13313104491218225635&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="scientific-writing">Scientific Writing</h4>
<ul>
<li><p><a
href="http://library.lol/main/8036CBB1CCC448CA7E036774D810EBC0">The uses
of argument</a> - <strong><em>Cambridge University Press</em></strong>,
1958. [<a
href="https://scholar.google.com/scholar?cluster=12052408655432810103&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Stephen Toulmin’s introduction to the Toulmin argument
pattern, which is generally consist of a claim, a justification, and a
rebuttal.</p></li>
<li><p><a href="https://www.jstor.org/stable/355200">A tagmemic approach
to paragraph analysis</a> - <strong><em>College Composition and
Communication</em></strong>, 1965. [<a
href="https://scholar.google.com/scholar?hl=en&amp;as_sdt=0%2C5&amp;q=A+Tagmemic+Approach+to+Paragraph+Analysis+AL+Becker&amp;btnG=">All
Versions</a>]. The original paper on analyzing the structure of
expository paragraphs, with the two patterns—the
Topic-Restriction-Illustration pattern and the Problem-Solution
pattern.</p></li>
<li><p><a
href="https://journals.sagepub.com/doi/abs/10.1177/0741088398015002004">The
uses and complexity of argument structures in expert and student
persuasive writing</a> - <strong><em>Written
Communication</em></strong>, 1998. [<a
href="https://scholar.google.com/scholar?cluster=3218190258774062869&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A behaviorial study revealing the argument structures
exploited by people in argumentative writing.</p></li>
<li><p><a
href="https://pure.mpg.de/rest/items/item_3020351/component/file_3045811/content">Towards
an argument interchange format</a> - <strong><em>The Knowledge
Engineering Review</em></strong>, 2006. [<a
href="https://scholar.google.com/scholar?cluster=11123720528835823517&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper introducing the Argument Interchange
Format (AIF) framework for argumentation analysis.</p></li>
<li><p><a
href="https://www.aaai.org/ocs/index.php/WS/AAAIW11/paper/viewFile/3940/4244">Speech
Acts of Argumentation: Inference Anchors and Peripheral Cues in
Dialogue</a> - <strong><em>AAAI’12</em></strong>, 2012. [<a
href="https://scholar.google.com/scholar?cluster=9761955212933152906&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper introducing the Information Anchoring
Theory (IAT) as an alternate for AIF.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="science-education">Science Education</h4>
<ul>
<li><p><a
href="https://www.harvardlds.org/wp-content/uploads/2018/05/Carey-Cognitive-science-and-science-education.-American-Psychologist.pdf">Cognitive
Science and Science Education</a> - <strong><em>American
Psychologist</em></strong>, 1986. [<a
href="https://scholar.google.com/scholar?cluster=6627805813997387166&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Susan Carey’s review on cognitive-science-based
methodologies for science education research.</p></li>
<li><p><a href="https://aclanthology.org/2023.acl-demo.2/">PersLEARN:
Research Training through the Lens of Perspective Cultivation</a> -
<strong><em>ACL’23</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=6242389165210232890">All
Versions</a>]. Scientific research is inherently shaped by its authors’
perspectives, influenced by various factors such as their personality,
community, or society. Junior researchers often face challenges in
identifying the perspectives reflected in the existing literature and
struggle to develop their own viewpoints. To address the problem, this
paper introduces PersLEARN, a tool designed to facilitate the
cultivation of scientific perspectives, starting from a basic seed idea
and progressing to a well-articulated framework.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="democratization-of-science">Democratization of Science</h4>
<ul>
<li><p><a
href="https://www.science.org/doi/full/10.1126/science.1250475">Reproducibility</a>
- <strong><em>Science</em></strong>, 2014. [<a
href="https://scholar.google.com/scholar?cluster=676974831306442279&amp;hl=en&amp;as_sdt=0,10">All
Versions</a>].</p></li>
<li><p><a
href="https://www.nature.com/articles/s41557-024-01470-8">Bridging the
information gap in organic chemical reactions</a> - <strong><em>Nature
Chemistry</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=5365091261196953334">All
Versions</a>]. This perspective article formulates eight principles to
improve data management in scientific publications relating to data
standardization, reproducibility and evaluation, and encourage
scientists to go beyond current publication standards.</p></li>
<li><p><a href="https://www.nature.com/articles/s41562-016-0021">A
manifesto for reproducible science</a> - <strong><em>Nature Human
Behavior</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=9515807942859203900&amp;hl=en&amp;as_sdt=0,10">All
Versions</a>].</p></li>
<li><p><a href="https://www.nature.com/articles/533452a">1,500
scientists lift the lid on reproducibility</a> -
<strong><em>Nature</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=11479406257389837824&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4204808/">How to Make
More Published Research True</a> - <strong><em>PLoS
Medicine</em></strong>, 2014. [<a
href="https://scholar.google.com/scholar?cluster=10945341175996677908">All
Versions</a>].</p></li>
<li><p><a href="https://www.nature.com/articles/d42473-019-00004-y">Six
factors affecting reproducibility in life science research and how to
handle them</a> - <strong><em>Nature
Advertisement</em></strong>.</p></li>
<li><p><a href="https://www.nature.com/articles/d41586-021-02428-3">Five
keys to writing a reproducible lab protocol</a> -
<strong><em>Nature</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=13259206850261301938">All
Versions</a>]. This interviewing paper introduces five ways to increase
the reproducibility of experimental protocols: (i) documenting protocols
as the experiment goes; (ii) providing video illustrations in addition
to written protocols; (iii) using electronic lab notebooks (ELNs) for
managing experimental resources digitally; (iv) depositing and
documenting reagents with understanding the rationale behind every step;
and (v) exploiting online platforms to share tips, extensions, methods,
and data among researchers.</p></li>
<li><p><a
href="https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.2003779">The
Experimental Design Assistant</a> - <strong><em>PLoS
Biology</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=12481490526120919925">All
Versions</a>]. [<a
href="https://www.nature.com/articles/nmeth.4462">Nature Methods
Correspondence</a>]. [<a href="https://eda.nc3rs.org.uk/">EDA
Website</a>]. The EDA is a web-based tool that guides the in vivo
researcher through the experimental design and analysis process,
providing automated feedback on the proposed design and generating a
graphical summary that aids communication with colleagues, funders,
regulatory authorities, and the wider scientific community.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="laboratory-automation">Laboratory Automation</h4>
<ul>
<li><p><a
href="https://www.science.org/doi/full/10.1126/science.aat0650">Reconfigurable
system for automated optimization of diverse chemical reactions</a> -
<strong><em>Science</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=3076614068291119943">All
Versions</a>]. [<a
href="https://www.science.org/doi/pdf/10.1126/science.aat0650">Preprint</a>].
This paper describes a plug-and-play, continuous-flow chemical synthesis
system that mitigates this challenge with an integrated combination of
hardware, software, and analytics. The system software controls the
user-selected reagents and unit operations (reactors and separators),
processes reaction analytics (high-performance liquid chromatography,
mass spectrometry, vibrational spectroscopy), and conducts automated
optimizations.</p></li>
<li><p><a
href="https://www.science.org/doi/full/10.1126/science.abc2986">A
universal system for digitization and automatic execution of the
chemical synthesis literature</a> - <strong><em>Science</em></strong>,
2020. [<a
href="https://scholar.google.com/scholar?cluster=13909991218383718512">All
Versions</a>]. [<a
href="https://www.chem.gla.ac.uk/cronin/images/pubs/Mehr-ScienceOct2020.pdf">Preprint</a>].
[<a href="https://croningroup.gitlab.io/chemputer/xdl/index.html">XDL
Documentation</a>]. [<a href="https://zenodo.org/records/3955107">XDL
Schema Database</a>]. This paper reports a software platform that uses
natural language processing to translate the organic chemistry
literature directly into editable code, which in turn can be compiled to
drive automated synthesis of the compound in the laboratory.</p></li>
<li><p><a
href="https://www.science.org/doi/full/10.1126/science.abo0058">Digitization
and validation of a chemical synthesis literature database in the
ChemPU</a> - <strong><em>Science</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=17368503277308594977">All
Versions</a>]. [<a
href="https://www.researchgate.net/profile/Aamir-Khan/publication/361857872_Digitization_and_validation_of_a_chemical_synthesis_literature_database_in_the_ChemPU/links/62cd356d00d0b451104cbfe9/Digitization-and-validation-of-a-chemical-synthesis-literature-database-in-the-ChemPU.pdf">Preprint</a>].
This paper presents an automatically executable chemical reaction
database of 100 molecules representative of the range of reactions found
in contemporary organic synthesis. The chemical reaction codes or χDLs
for the reactions have been stored in a database for version control,
validation, collaboration, and data mining. Of these syntheses, more
than 50 entries from the database have been downloaded and robotically
run in seven modular chemputers with yields and purities comparable to
those achieved by an expert chemist.</p></li>
<li><p><a
href="https://pubs.acs.org/doi/full/10.1021/jacsau.1c00303">Chemputation
and the Standardization of Chemical Informatics</a> -
<strong><em>Journal of the American Chemical Society (Au)</em></strong>,
2021. [<a
href="https://scholar.google.com/scholar?cluster=3884902150148113559">All
Versions</a>]. This paper describes a standard hardware (the chemical
processing programming architecture — the ChemPU) to encompass all
chemical synthesis, an approach which unifies all chemistry automation
strategies, from solid-phase peptide synthesis, to HTE flow chemistry
platforms, while at the same time establishing a publication standard so
that researchers can exchange chemical code (χDL) to ensure
reproducibility and interoperability.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41557-020-00596-9">Convergence of
multiple synthetic paradigms in a universally programmable chemical
synthesis machine</a> - <strong><em>Nature Chemistry</em></strong>,
2020. [<a
href="https://scholar.google.com/scholar?cluster=18024303106901939347">All
Versions</a>]. [<a
href="https://eprints.gla.ac.uk/231947/">Preprint</a>]. This paper shows
how the Chemputer synthesis robot can be programmed to perform many
different reactions, including solid-phase peptide synthesis, iterative
cross-coupling and accessing reactive, unstable diazirines in a single,
unified system with high yields and purity.</p></li>
<li><p><a href="https://www.nature.com/articles/s41557-022-01016-w">An
autonomous portable platform for universal chemical synthesis</a> -
<strong><em>Nature Chemistry</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=4484997534431409967">All
Versions</a>]. [<a
href="https://eprints.gla.ac.uk/275574/">Preprint</a>]. This paper
presents a portable suitcase-sized chemical synthesis platform
containing all the modules required for synthesis and purification. The
system uses a chemical programming language coupled to a digital reactor
generator to produce reactors and executable protocols based on
text-based literature syntheses. Simultaneously, the platform generates
a reaction pressure fingerprint, used to monitor processes within the
reactors and remotely perform a protocol quality control.</p></li>
<li><p><a href="https://www.nature.com/articles/s41467-024-45444-3">An
integrated self-optimizing programmable chemical synthesis and reaction
engine</a> - <strong><em>Nature Communications</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=9157508627971047184">All
Versions</a>]. This paper presents a dynamically programmable system
capable of making, optimizing, and discovering new molecules which
utilizes seven sensors that continuously monitor the reaction. By
developing a dynamic programming language, the work demonstrates the
10-fold scale-up of a highly exothermic oxidation reaction, end point
detection, as well as detecting critical hardware failures.</p></li>
<li><p><a href="https://www.nature.com/articles/s41586-020-2442-2">A
mobile robotic chemist</a> - <strong><em>Nature</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=13216902493789027324&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://strathprints.strath.ac.uk/74759/1/Burger_etal_Nature_2020_A_mobile_robotic.pdf">Preprint</a>].
This work uses a mobile robot to search for improved photocatalysts for
hydrogen production from water. The robot operated autonomously over
eight days, performing 688 experiments within a ten-variable
experimental space, driven by a batched Bayesian search algorithm. This
autonomous search identified photocatalyst mixtures that were six times
more active than the initial formulations, selecting beneficial
components and deselecting negative ones.</p></li>
<li><p><a href="https://www.nature.com/articles/s41586-023-06734-w">An
autonomous laboratory for the accelerated synthesis of novel
materials</a> - <strong><em>Nature</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=17944003281308189532">All
Versions</a>]. This paper introduces the A-Lab, an autonomous laboratory
for the solid-state synthesis of inorganic powders. This platform uses
computations, historical data from the literature, machine learning (ML)
and active learning to plan and interpret the outcomes of experiments
performed using robotics. Over 17 days of continuous operation, the
A-Lab realized 41 novel compounds from a set of 58 targets including a
variety of oxides and phosphates that were identified using large-scale
ab initio phase-stability data from the Materials Project and Google
DeepMind.</p></li>
<li><p><a href="https://www.nature.com/articles/542125a">The Internet of
Things comes to the lab</a> - <strong><em>Nature</em></strong>, 2017.
[<a
href="https://scholar.google.com/scholar?cluster=7747117198956166976&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The emergence of connected instruments and equipment
promises to untether researchers from the laboratory — letting them
fine-tune experiments and analyse data remotely.</p></li>
<li><p><a href="https://www.nature.com/articles/s41467-023-44599-9">A
dynamic knowledge graph approach to distributed self-driving
laboratories</a> - <strong><em>Nature Communications</em></strong>,
2024. [<a
href="https://scholar.google.com/scholar?cluster=7070798385652764751">All
Versions</a>]. This work employs ontologies to capture data and material
flows in design-make-test-analyse cycles, utilising autonomous agents as
executable knowledge components to carry out the experimentation
workflow. Data provenance is recorded to ensure its findability,
accessibility, interoperability, and reusability. The architecture is
built upon the World Avatar project, which seeks to create an
all-encompassing digital twin based on a dynamic knowledge
graph.</p></li>
<li><p><a
href="https://pubs.rsc.org/en/content/articlehtml/2021/sc/d1sc04588a">Automation
isn’t automatic</a> - <strong><em>Chemical Science</em></strong>, 2021.
[<a
href="https://scholar.google.com/scholar?cluster=14176714971050097971">All
Versions</a>]. This perspective provides an overview of the current
state of automation of synthetic chemistry at the benchtop scale with a
particular emphasis on core considerations and the ensuing challenges of
deploying a system. The authors aim to reframe automation as decidedly
not automatic but rather an iterative process that involves a series of
careful decisions (both human and computational) and constant
adjustment.</p></li>
<li><p><a
href="https://www.cell.com/trends/chemistry/fulltext/S2589-5974(23)00249-6">Balancing
act: when to flex and when to stay fixed</a> - <strong><em>Trends in
Chemistry</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=14208571639305934551">All
Versions</a>]. This perspective article provides essential insights into
the decision-making process for choosing automation platforms,
highlighting the suitability of fixed automation for standardized tasks
and the strategic use of flexible automation in dynamic research
settings.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S2590238522006385">What
is a minimal working example for a self-driving laboratory?</a> -
<strong><em>Matter</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=1612804023616680548">All
Versions</a>]. This paper proposes SDL-Demo: a low-cost “Hello, World!”
for self-driving laboratories that combines “Hello, World!” tasks from
electronics, physics-based simulations, and optimization. SDL-Demo is
modular and extensible, making it an ideal candidate for low-cost
teaching and prototyping of self-driving laboratory concepts.</p></li>
<li><p><a href="https://elifesciences.org/articles/77007">Robotic search
for optimal cell culture in regenerative medicine</a> -
<strong><em>eLife</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=1330075145723138159">All
Versions</a>]. This paper develops a robotic AI system with a batch
Bayesian optimization algorithm that autonomously induces the
differentiation of induced pluripotent stem cell-derived retinal pigment
epithelial (iPSC-RPE) cells. From 200 million possible parameter
combinations, the system performed cell culture in 143 different
conditions in 111 days, resulting in 88% better iPSC-RPE production than
that obtained by the pre-optimized culture in terms of the pigmentation
scores.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="ai-assisted-research">AI Assisted Research</h4>
<ul>
<li><p><a
href="https://www.nature.com/articles/s41586-023-06221-2">Scientific
discovery in the age of artificial intelligence</a> -
<strong><em>Nature</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=11962817646389491592">All
Versions</a>]. A review article that examines breakthroughs over the
past decade that include self-supervised learning, which allows models
to be trained on vast amounts of unlabelled data, and geometric deep
learning, which leverages knowledge about the structure of scientific
data to enhance model accuracy and efficiency.</p></li>
<li><p><a href="https://arxiv.org/abs/2311.07361">The Impact of Large
Language Models on Scientific Discovery: a Preliminary Study using
GPT-4</a> - <strong><em>Microsoft Research AI4Science</em></strong>,
2023. [<a href="">All Versions</a>]. [<a
href="https://github.com/microsoft/LLM4ScientificDiscovery">Project</a>].
A survey on the performance of LLMs within the context of scientific
discovery, focusing on GPT-4.</p></li>
<li><p><a
href="https://www.science.org/doi/full/10.1126/sciadv.aay4275">Machine
learning-assisted molecular design and efficiency prediction for
high-performance organic photovoltaic materials</a> -
<strong><em>Science Advances</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=12392230644945701722">All
Versions</a>]. In the process of finding high-performance materials for
organic photovoltaics (OPVs), it is meaningful if one can establish the
relationship between chemical structures and photovoltaic properties
even before synthesizing them. This work first establishes a database
containing over 1700 donor materials reported in the literature. Through
supervised learning, our machine learning (ML) models can build up the
structure-property relationship and, thus, implement fast screening of
OPV materials. The authors explore several expressions for molecule
structures, i.e., images, ASCII strings, descriptors, and fingerprints,
as inputs for various ML algorithms. It is found that fingerprints with
length over 1000 bits can obtain high prediction accuracy. The
reliability of the approach is further verified by screening 10 newly
designed donor materials. Good consistency between model predictions and
experimental outcomes is obtained. The result indicates that ML is a
powerful tool to prescreen new OPV materials, thus accelerating the
development of the OPV field.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41598-018-34533-1">Design of
metalloproteins and novel protein folds using variational
autoencoders</a> - <strong><em>Scientific Reports</em></strong>, 2018.
[<a
href="https://scholar.google.com/scholar?cluster=18126187509308242959">All
Versions</a>]. The design of novel proteins has many applications but
remains an attritional process with success in isolated cases.
Meanwhile, deep learning technologies have exploded in popularity in
recent years and are increasingly applicable to biology due to the rise
in available data. This work attempts to link protein design and deep
learning by using variational autoencoders to generate protein sequences
conditioned on desired properties. Potential copper and calcium binding
sites are added to non-metal binding proteins without human intervention
and compared to a hidden Markov model. In another use case, a grammar of
protein structures is developed and used to produce sequences for a
novel protein topology. One candidate structure is found to be stable by
molecular dynamics simulation. The ability of the model to confine the
vast search space of protein sequences and to scale easily has the
potential to assist in a variety of protein design tasks.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41586-021-03819-2">Highly
accurate protein structure prediction with AlphaFold</a> -
<strong><em>Nature</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=6286436358625670901">All
Versions</a>]. This paper provides the first computational method that
can regularly predict protein structures with atomic accuracy even in
cases in which no similar structure is known. This approach is a
canonical application of observation- and explanation- based method for
protein structure prediction instead of first-principle-based
methods.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41586-023-05773-7">Human–machine
collaboration for improving semiconductor process development</a> -
<strong><em>Nature</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=10295771969614897767">All
Versions</a>]. [<a
href="https://www.nature.com/articles/d41586-023-01353-x">Nature
News</a>]. This work studies Bayesian optimization algorithms to
investigate how artificial intelligence (AI) might decrease the cost of
developing complex semiconductor chip processes. In particular, this
work create a controlled virtual process game to systematically
benchmark the performance of humans and computers for the design of a
semiconductor fabrication process. The authors find that human engineers
excel in the early stages of development, whereas the algorithms are far
more cost-efficient near the tight tolerances of the target.</p></li>
<li><p><a href="https://www.nature.com/articles/s41586-023-06555-x">A
foundation model for generalizable disease detection from retinal
images</a> - <strong><em>Nature</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=3139988207343394501">All
Versions</a>]. This paper presents RETFound, a foundation model for
retinal images that learns generalizable representations from unlabelled
retinal images and provides a basis for label-efficient model adaptation
in several applications. Specifically, RETFound is trained on
1.6 million unlabelled retinal images by means of self-supervised
learning and then adapted to disease detection tasks with explicit
labels.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41586-023-06185-3">Accurate
medium-range global weather forecasting with 3D neural networks</a> -
<strong><em>Nature</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=7198604620204619820">All
Versions</a>]. This paer introduces an artificial-intelligence-based
method for accurate, medium-range global weather forecasting. It shows
that three-dimensional deep networks equipped with Earth-specific priors
are effective at dealing with complex patterns in weather data, and that
a hierarchical temporal aggregation strategy reduces accumulation errors
in medium-range forecasting. Trained on 39 years of global data, the
program, Pangu-Weather, obtains stronger deterministic forecast results
on reanalysis data in all tested variables when compared with the
world’s best NWP system, the operational integrated forecasting system
of the European Centre for Medium-Range Weather Forecasts.</p></li>
<li><p><a
href="https://www.science.org/doi/10.1126/science.adi2336">Learning
skillful medium-range global weather forecasting</a> -
<strong><em>Science</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=269756601245477923&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.nature.com/articles/s41586-023-06184-4">Skilful
nowcasting of extreme precipitation with NowcastNet</a> -
<strong><em>Nature</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=17837864391812838009&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.nature.com/articles/s41586-023-06792-0">Autonomous
chemical research with large language models</a> -
<strong><em>Nature</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=8097577445064259203">All
Versions</a>]. An artificial intelligence system driven by GPT-4 that
autonomously designs, plans and performs complex experiments by
incorporating large language models empowered by tools such as internet
and documentation search, code execution and experimental
automation.</p></li>
<li><p><a
href="https://www.nature.com/articles/s42256-024-00832-8">Augmenting
large language models with chemistry tools</a> - <strong><em>Nature
Machine Intelligence</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=9291969834799338362">All
Versions</a>]. [<a
href="https://arxiv.org/abs/2304.05376">Preprint</a>]. This paper
introduces ChemCrow, an LLM chemistry agent designed to accomplish tasks
across organic synthesis, drug discovery and materials design. By
integrating 18 expert-designed tools and using GPT-4 as the LLM,
ChemCrow augments the LLM performance in chemistry, and new capabilities
emerge. The agent autonomously planned and executed the syntheses of an
insect repellent and three organocatalysts and guided the discovery of a
novel chromophore.</p></li>
<li><p><a
href="https://aclanthology.org/2023.emnlp-main.162/">BioPlanner:
Automatic Evaluation of LLMs on Protocol Planning in Biology</a> -
<strong><em>EMNLP’23</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=1222312709622462659">All
Versions</a>]. [<a
href="https://github.com/bioplanner/bioplanner">Project</a>]. This paper
presents an automatic evaluation framework for the task of planning
experimental protocols, and introduces BioProt: a dataset of biology
protocols with corresponding pseudocode representations.</p></li>
<li><p><a href="https://www.nature.com/articles/s41467-024-47997-9">A
human-machine interface for automatic exploration of chemical reaction
networks</a> - <strong><em>Nature Communications</em></strong>, 2024.
[<a
href="https://scholar.google.com/scholar?cluster=13306522324804014261">All
Versions</a>]. Autonomous reaction network exploration algorithms offer
a systematic approach to explore mechanisms of complex chemical
processes. However, the resulting reaction networks are so vast that an
exploration of all potentially accessible intermediates is
computationally too demanding. This paper introduces a STEERING WHEEL to
guide an otherwise unbiased automated exploration. The STEERING WHEEL
algorithm is intuitive, generally applicable, and enables one to focus
on specific regions of an emerging network. It also allows for guiding
automated data generation in the context of mechanism exploration,
catalyst design, and other chemical optimization challenges.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41467-024-50779-y">PatCID: an
open-access dataset of chemical structures in patent documents</a> -
<strong><em>Nature Communications</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=329287456191953845">All
Versions</a>]. The automatic analysis of patent publications has
potential to accelerate research across various domains, including drug
discovery and material science. Within patent documents, crucial
information often resides in visual depictions of molecule structures.
PatCID (Patent-extracted Chemical-structure Images database for
Discovery) allows to access such information at scale. It enables users
to search which molecules are displayed in which documents. PatCID
contains 81M chemical-structure images and 14M unique chemical
structures. This work compares PatCID with state-of-the-art chemical
patent-databases. On a random set, PatCID retrieves 56.0% of molecules,
which is higher than automatically-created databases, Google Patents
(41.5%) and SureChEMBL (23.5%), as well as manually-created databases,
Reaxys (53.5%) and SciFinder (49.5%). Leveraging state-of-the-art
methods of document understanding, PatCID high-quality data outperforms
currently available automatically-generated patent-databases. PatCID
even competes with proprietary manually-created patent-databases. This
enables promising applications for automatic literature review and
learning-based molecular generation methods.</p></li>
<li><p><a href="https://arxiv.org/abs/2311.00176">ChipNeMo:
Domain-Adapted LLMs for Chip Design</a> - 2023. [<a
href="https://scholar.google.com/scholar?cluster=5962372610489019326">All
Versions</a>]. ChipNeMo aims to explore the applications of large
language models (LLMs) for industrial chip design. Instead of directly
deploying off-the-shelf commercial or open-source LLMs, the authors
instead adopt the following domain adaptation techniques:
domain-adaptive tokenization, domain-adaptive continued pretraining,
model alignment with domain-specific instructions, and domain-adapted
retrieval models. The authors evaluate these methods on three selected
LLM applications for chip design: an engineering assistant chatbot, EDA
script generation, and bug summarization and analysis. Evaluations
demonstrate that domain-adaptive pretraining of language models, can
lead to superior performance in domain related downstream tasks compared
to their base LLaMA2 counterparts, without degradations in generic
capabilities. In particular, the largest model, ChipNeMo-70B,
outperforms the highly capable GPT-4 on two of the use cases, namely
engineering assistant chatbot and EDA scripts generation, while
exhibiting competitive performance on bug summarization and analysis.
These results underscore the potential of domain-specific customization
for enhancing the effectiveness of large language models in specialized
applications.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41467-021-22048-9">Single-atom
alloy catalysts designed by first-principles calculations and artificial
intelligence</a> - <strong><em>Nature Communications</em></strong>,
2021. [<a
href="https://scholar.google.com/scholar?cluster=6593978922251447907">All
Versions</a>]. This paper addresses the problem of new Single-atom-alloy
catalysts (SAACs) discovery by applying a compressed-sensing
data-analytics approach parameterized with density-functional
inputs.</p></li>
<li><p><a
href="https://www.pnas.org/doi/abs/10.1073/pnas.2016239118">Biological
structure and function emerge from scaling unsupervised learning to 250
million protein sequences</a> - <strong><em>Proceedings of the National
Academy of Sciences</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=15181490380139888639&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007/s00449-016-1659-9">Comparability
of automated human induced pluripotent stem cell culture: a pilot
study</a> - <strong><em>Bioprocess and Biosystems
Engineering</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=14666375402220991095&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://pubs.acs.org/doi/full/10.1021/jacs.4c00338">Artificial
Intelligence for Retrosynthetic Planning Needs Both Data and Expert
Knowledge</a> - <strong><em>Journal of the American Chemical
Society</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=10595951443492961310">All
Versions</a>]. The development of AI synthesis planners trained solely
on reaction-example-data has stagnated and is not on par with the
performance of “hybrid” algorithms combining AI with expert knowledge.
This Perspective examines possible causes of these shortcomings,
extending beyond the established reasoning of insufficient quantities of
reaction data. Drawing attention to the intricacies and data biases that
are specific to the domain of synthetic chemistry, the authors advocate
augmenting the unique capabilities of AI with the knowledge base and the
reasoning strategies of domain experts. By actively involving synthetic
chemists, who are the end users of any synthesis planning software, into
the development process, the authors envision to bridge the gap between
computer algorithms and the intricate nature of chemical
synthesis.</p></li>
<li><p><a
href="https://www.cell.com/cell-reports-medicine/fulltext/S2666-3791(21)00197-X">Virtual
and augmented reality for biomedical applications</a> - <strong><em>Cell
Reports Medicine</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=14732259085495422063">All
Versions</a>]. 3D visualization technologies such as virtual reality
(VR), augmented reality (AR), and mixed reality (MR) have gained
popularity in the recent decade. Digital extended reality (XR)
technologies have been adopted in various domains ranging from
entertainment to education because of their accessibility and
affordability. XR modalities create an immersive experience, enabling 3D
visualization of the content without a conventional 2D display
constraint. This paper provides a perspective on XR in current
biomedical applications and demonstrate case studies using cell biology
concepts, multiplexed proteomics images, surgical data for heart
operations, and cardiac 3D models. Emerging challenges associated with
XR technologies in the context of adverse health effects and a cost
comparison of distinct platforms are discussed. The presented XR
platforms will be useful for biomedical education, medical training,
surgical guidance, and molecular data visualization to enhance trainees’
and students’ learning, medical operation accuracy, and the
comprehensibility of complex biological systems.</p></li>
<li><p><a href="https://www.nature.com/articles/s41591-019-0539-7">An
augmented reality microscope with real-time artificial intelligence
integration for cancer diagnosis</a> - <strong><em>Nature
Medicine</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=3280260879383275625">All
Versions</a>]. The microscopic assessment of tissue samples is
instrumental for the diagnosis and staging of cancer, and thus guides
therapy. However, these assessments demonstrate considerable variability
and many regions of the world lack access to trained pathologists.
Though artificial intelligence (AI) promises to improve the access and
quality of healthcare, the costs of image digitization in pathology and
difficulties in deploying AI solutions remain as barriers to real-world
use. This work proposes a cost-effective solution: the augmented reality
microscope (ARM). The ARM overlays AI-based information onto the current
view of the sample in real time, enabling seamless integration of AI
into routine workflows.</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/abstract/document/10059206">Optimizing
Spaced Repetition Schedule by Capturing the Dynamics of Memory</a> -
<strong><em>IEEE Transactions on Knowledge and Data
Engineering</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=949715967083833369&amp;hl=en&amp;as_sdt=0,10">All
Versions</a>].</p></li>
<li><p><a
href="https://aclanthology.org/2020.findings-emnlp.261/">LEGAL-BERT: The
Muppets straight out of Law School</a> -
<strong><em>EMNLP’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=11254432523766039890&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Generating answers to legal questions, analyze contracts,
and summarizing legal documents, making legal knowledge more accessible
to non-experts.</p></li>
<li><p><a
href="https://academic.oup.com/bioinformatics/article/36/4/1234/5566506">BioBERT:
a pre-trained biomedical language representation model for biomedical
text mining</a> - <strong><em>Bioinformatics</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=2783127196632783403&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Answering medical questions, identifying relevant
clinical trials, and diagnosing diseases based on symptoms, making
medical information more accessible to the general public.</p></li>
<li><p><a
href="https://dl.acm.org/doi/abs/10.5555/3491440.3492062">Finbert: A
pre-trained financial language representation model for financial text
mining</a> - <strong><em>IJCAI’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=17844713837232165872&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Predicting stock market trends, analyzing financial
documents, and generating summaries of economic news articles, helping
to disseminate financial knowledge.</p></li>
<li><p><a href="https://aclanthology.org/D19-1371/">SciBERT: A
Pretrained Language Model for Scientific Text</a> -
<strong><em>EMNLP’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=7377999893003631695&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Searching and synthesizing scientific literature, aiding
researchers in hypothesis generation, and assisting with experimental
design, making scientific knowledge more accessible.</p></li>
<li><p><a
href="https://aclanthology.org/2020.findings-emnlp.139/">CodeBERT: A
Pre-Trained Model for Programming and Natural Languages</a> -
<strong><em>EMNLP’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=9055786889913621082&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Completing code, generating programming documentation,
and providing technical support, making programming knowledge more
accessible to non-experts.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="theory-of-mind">Theory of Mind</h3>
<ul>
<li><p><a href="https://en.wikipedia.org/wiki/Theory_of_mind">Theory of
Mind</a> - <strong><em>Wikipedia</em></strong>. Wikipedia on Theory of
Mind (ToM), a cognitive capability that estimating others’ goal, belief,
and desire.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/intentionality/">Intentionality</a>
- <strong><em>Plato Stanford</em></strong>.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/mental-imagery/">Mental
Imagery</a> - <strong><em>Plato Stanford</em></strong>.</p></li>
</ul>
<!--* [Cognitive Science](https://plato.stanford.edu/entries/cognitive-science/) - ***Plato Stanford***.
* [The Mind/Brain Identity Theory](https://plato.stanford.edu/entries/mind-identity/) - ***Plato Stanford***.
* [Mental Representation](https://plato.stanford.edu/entries/mental-representation/) - ***Plato Stanford***.
* [Temporal Consciousness](https://plato.stanford.edu/entries/consciousness-temporal/) - ***Plato Stanford***.
* [The Experience and Perception of Time](https://plato.stanford.edu/entries/time-experience/) - ***Plato Stanford***.
* [Practical Reason](https://plato.stanford.edu/entries/practical-reason/) - ***Plato Stanford***.
* [Memory](https://plato.stanford.edu/entries/memory/) - ***Plato Stanford***.-->
<!-- * [The Computational Theory of Mind](https://plato.stanford.edu/entries/computational-mind/) - ***Plato Stanford***. A computational philosophy account on ToM. -->
<ul>
<li><p><a
href="https://www.cell.com/trends/cognitive-sciences/fulltext/S1364-6613(16)30053-5">The
naïve utility calculus: Computational principles underlying commonsense
psychology</a> - <strong><em>Trends in Cognitive Sciences</em></strong>,
2016. [<a
href="https://scholar.google.com/scholar?cluster=6894095575934067763">All
Versions</a>]. [<a
href="http://sll.stanford.edu/docs/2016_JaraEttinger_Gweon_Schulz_Tenenbaum_TiCS.pdf">Preprint</a>].
This review article proposes that human social cognition is structured
around a basic understanding of ourselves and others as intuitive
utility maximizers: from a young age, humans implicitly assume that
agents choose goals and actions to maximize the rewards they expect to
obtain relative to the costs they expect to incur. This ‘naïve utility
calculus’ allows both children and adults observe the behavior of others
and infer their beliefs and desires, their longer-term knowledge and
preferences, and even their character: who is knowledgeable or
competent, who is praiseworthy or blameworthy, who is friendly,
indifferent, or an enemy.</p></li>
<li><p><a
href="https://www.cell.com/trends/cognitive-sciences/fulltext/S1364-6613(22)00185-1">Planning
with theory of mind</a> - <strong><em>Trends in Cognitive
Sciences</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=8461125353366208047">All
Versions</a>]. [<a
href="https://saxelab.mit.edu/sites/default/files/publications/HoSaxeCushman2022.pdf">Preprint</a>].
A perspective on understanding Theory of Mind through planning that
consists of abstract structured causal representations and supports
efficient search and selection from innumerable possible actions.
Planning requires that Theory of Mind consists of abstract structured
causal representations and supports efficient search and selection from
innumerable possible actions. Theory of Mind contrasts with less
cognitively demanding alternatives: statistical predictive models of
other people’s actions, or model-free reinforcement of actions by their
effects on other people. Theory of Mind is likely used to plan novel
interventions and predict their effects, for example, in pedagogy,
emotion regulation, and impression management.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0010027709001607">Action
Understanding as Inverse Planning</a> -
<strong><em>Cognition</em></strong>, 2009. [<a
href="https://scholar.google.com/scholar?cluster=11478704181983566675">All
Versions</a>]. [<a
href="https://ars.els-cdn.com/content/image/1-s2.0-S0010027709001607-mmc1.pdf">Appendix</a>].
The original paper on Inverse Planning, a computational implementation
of Theory of Mind. Humans are adept at inferring the mental states
underlying other agents’ actions, such as goals, beliefs, desires,
emotions and other thoughts. This paper proposes a computational
framework based on Bayesian inverse planning for modeling human action
understanding. The framework represents an intuitive theory of
intentional agents’ behavior based on the principle of rationality: the
expectation that agents will plan approximately rationally to achieve
their goals, given their beliefs about the world. The mental states that
caused an agent’s behavior are inferred by inverting this model of
rational planning using Bayesian inference, integrating the likelihood
of the observed actions with the prior over mental states.</p></li>
<li><p><a
href="https://escholarship.org/content/qt5rk7z59q/qt5rk7z59q.pdf">Bayesian
Theory of Mind: Modeling Joint Belief-Desire Attribution</a> -
<strong><em>CogSci’11</em></strong>, 2011. [<a
href="https://scholar.google.com/scholar?cluster=7454981153033683025">All
Versions</a>]. [<a
href="http://web.mit.edu/9.s915/www/classes/theoryOfMind.pdf">Preprint</a>].
This paper presents a computational framework for understanding Theory
of Mind (ToM): the human capacity for reasoning about agents’ mental
states such as beliefs and desires. The proposed Bayesian model of ToM
(or BToM) expresses the predictive model of belief- and desire-dependent
action at the heart of ToM as a partially observable Markov decision
process (POMDP), and reconstructs an agent’s joint belief state and
reward function using Bayesian inference, conditioned on observations of
the agent’s behavior in some environmental context.</p></li>
<li><p><a href="https://psyarxiv.com/f692k/">The Signature of All
Things: Children Infer Knowledge States from Static Images</a> -
<strong><em>CogSci’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=12380982112592086477&amp;hl=en&amp;as_sdt=0,5&amp;as_ylo=2017">All
Versions</a>].</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S1364661316301565?via%3Dihub">Bayesian
Brains without Probabilities</a> - <strong><em>Trends in Cognitive
Sciences</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=13076510377612067772&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A perspective on human probabilistic modeling without
explicit probabilistic computation.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41562-017-0064">Rational
quantitative attribution of beliefs, desires and percepts in human
mentalizing</a> - <strong><em>Nature Human Behavior</em></strong>, 2017.
[<a
href="https://scholar.google.com/scholar?cluster=9377509910551057835">All
Versions</a>]. [<a
href="https://cbmm.mit.edu/sites/default/files/publications/article.pdf">Preprint</a>].
This paper presents a model of core mentalizing computations: inferring
jointly an actor’s beliefs, desires and percepts from how they move in
the local spatial environment. The proposed Bayesian theory of mind
(BToM) model is based on probabilistically inverting
artificial-intelligence approaches to rational planning and state
estimation, which extend classical expected-utility agent models to
sequential actions in complex, partially observable domains.</p></li>
<li><p><a
href="http://proceedings.mlr.press/v80/rabinowitz18a.html">Machine
theory of mind</a> - <strong><em>ICML’18</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=6267278380616425333">All
Versions</a>]. Theory of mind (ToM) broadly refers to humans’ ability to
represent the mental states of others, including their desires, beliefs,
and intentions. This work proposes a Theory of Mind neural network — a
ToMnet — which uses meta-learning to build such models of the agents it
encounters. The ToMnet learns a strong prior model for agents’ future
behaviour, and, using only a small number of behavioural observations,
can bootstrap to richer predictions about agents’ characteristics and
mental states.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S2352154618302055?via%3Dihub">Theory
of mind as inverse reinforcement learning</a> - <strong><em>Current
Opinion in Behavioral Sciences</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=14959443239271810913">All
Versions</a>]. This paper reviews the idea that Theory of Mind — humans’
ability to reason about other people’s mental states — can be formalized
as inverse reinforcement learning. Under this framework, expectations
about how mental states produce behavior are captured in a reinforcement
learning (RL) model. Predicting other people’s actions is achieved by
simulating a RL model with the hypothesized beliefs and desires, while
mental-state inference is achieved by inverting this model. Although
many advances in inverse reinforcement learning (IRL) did not have human
Theory of Mind in mind, this paper focuses on what they reveal when
conceptualized as cognitive theories.</p></li>
<li><p><a
href="https://onlinelibrary.wiley.com/doi/full/10.1111/tops.12371">Computational
Models of Emotion Inference in Theory of Mind: A Review and Roadmap</a>
- <strong><em>Topics in Cognitive Science</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=15919410726494658168">All
Versions</a>]. This paper proposes an intuitive theory framework to
studying affective cognition—how humans reason about emotions—and derive
a taxonomy of inferences within affective cognition. Using this
taxonomy, the authors review formal computational modeling work on such
inferences, including causal reasoning about how others react to events,
reasoning about unseen causes of emotions, reasoning with multiple cues,
as well as reasoning from emotions to other mental states. This
framework proposes unifying these various types of reasoning as Bayesian
inference within a common “intuitive Theory of Emotion.”</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0010028520300633">The
Naïve Utility Calculus as a unified, quantitative framework for action
understanding</a> - <strong><em>Cognitive Psychology</em></strong>,
2021. [<a
href="https://scholar.google.com/scholar?cluster=10366690800692546587">All
Versions</a>]. [<a
href="http://www.github.com/julianje/bishop">Project</a>]. This paper
presents a formal theory of the Naïve Utility Calculus as a
probabilistic generative model, which highlights the role of cost and
reward tradeoffs in a Bayesian framework for action-understanding. The
model predicts with quantitative accuracy how people infer agents’
subjective costs and rewards based on their observable actions. By
distinguishing between desires, goals, and intentions, the model extends
to complex action scenarios unfolding over space and time in scenes with
multiple objects and multiple action episodes.</p></li>
<li><p><a href="http://proceedings.mlr.press/v139/shu21a.html">AGENT: A
Benchmark for Core Psychological Reasoning</a> -
<strong><em>ICML’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=9729067071974484204">All
Versions</a>]. Inspired by cognitive development studies on intuitive
psychology, this paper presents a benchmark consisting of a large
dataset of procedurally generated 3D animations, AGENT (Action, Goal,
Efficiency, coNstraint, uTility), structured around four scenarios (goal
preferences, action efficiency, unobserved constraints, and cost-reward
trade-offs) that probe key concepts of core intuitive psychology. The
results suggest that to pass the designed tests of core intuitive
psychology at human levels, a model must acquire or have built-in
representations of how agents plan, combining utility computations and
core knowledge of objects and physics.</p></li>
<li><p><a
href="https://www.annualreviews.org/doi/pdf/10.1146/annurev-psych-081420-110718">Experimental
Games and Social Decision Making</a> - <strong><em>Annual Review of
Psychology</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=4713510112126264116">All
Versions</a>]. Experimental games model situations in which the future
outcomes of individuals and groups depend on their own choices and on
those of other (groups of) individuals. Games are a powerful tool to
identify the neural and psychological mechanisms underlying
interpersonal and group cooperation and coordination. This review
article discusses recent developments in how experimental games are used
and adapted, with an increased focus on repeated interactions, partner
control through sanctioning, and partner (de)selection for future
interactions.</p></li>
<li><p><a
href="https://www.aaai.org/ojs/index.php/AAAI/article/view/4574">Theory
of Minds: Understanding Behavior in Groups through Inverse Planning</a>
- <strong><em>AAAI’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=6755247312077985817">All
Versions</a>]. Towards the goal of building machine-learning algorithms
with human-like social intelligence, this paper develops a generative
model of multiagent action understanding based on a novel representation
for these latent relationships called Composable Team Hierarchies (CTH).
This representation is grounded in the formalism of stochastic games and
multi-agent reinforcement learning. This work uses CTH as a target for
Bayesian inference yielding a new algorithm for understanding behavior
in groups that can both infer hidden relationships as well as predict
future actions for multiple agents interacting together.</p></li>
<li><p><a
href="https://psycnet.apa.org/fulltext/2019-58384-001.pdf?auth_token=0859666184839448b848053cd7bdceb2bdf2745a">Leveraging
Facial Expressions and Contextual Information to Investigate Opaque
Representations of Emotion</a> - <strong><em>Emotion</em></strong>,
2019. [<a
href="https://scholar.google.com/scholar?cluster=9634378462684744548&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://linkinghub.elsevier.com/retrieve/pii/S0010027712002235">Waiting
and weighting: Information sampling is a balance between efficiency and
error-reduction</a> - <strong><em>Cognition</em></strong>, 2013. [<a
href="https://scholar.google.com/scholar?cluster=12787722822882067638&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0896627313005503?via%3Dihub">Natural
scene statistics account for the representation of scene categories in
human visual cortex</a> - <strong><em>Neuron</em></strong>, 2013. [<a
href="https://scholar.google.com/scholar?cluster=14030885492052338412&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.nature.com/articles/s41598-018-23618-6">Using human
brain activity to guide machine learning</a> - <strong><em>Scientific
Report</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=12987955253653036948&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://psycnet.apa.org/record/2019-27729-001">Unit of
visual working memory: A Boolean map provides a better account than an
object does</a> - <strong><em>Journal of Experimental
Psychology</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=14909735035752892020&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://www.pnas.org/content/117/42/26158.short">The
logic of universalization guides moral judgment</a> -
<strong><em>Proceedings of the National Academy of
Sciences</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=13482051983012049752&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://openaccess.thecvf.com/content/CVPR2021/html/Fan_Learning_Triadic_Belief_Dynamics_in_Nonverbal_Communication_From_Videos_CVPR_2021_paper.html">Learning
Triadic Belief Dynamics in Nonverbal Communication from Videos</a> -
<strong><em>CVPR’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=15365483338824697316">All
Versions</a>]. [<a
href="https://arxiv.org/abs/2104.02841">Preprint</a>]. This paper
incorporates different nonverbal communication cues (e.g., gaze, human
poses, and gestures) to represent, model, learn, and infer agents’
mental states from pure visual inputs. Crucially, such a mental
representation takes the agent’s belief into account so that it
represents what the true world state is and infers the beliefs in each
agent’s mental state, which may differ from the true world states. By
aggregating different beliefs and true world states, the model
essentially forms “five minds” during the interactions between two
agents. This “five minds” model differs from prior works that infer
beliefs in an infinite recursion; instead, agents’ beliefs are converged
into a “common mind”. Based on this representation, this work further
devises a hierarchical energy-based model that jointly tracks and
predicts all five minds. From this new perspective, a social event is
interpreted by a series of nonverbal communication and belief dynamics,
which transcends the classic keyframe video summary.</p></li>
<li><p><a
href="https://dspace.mit.edu/bitstream/handle/1721.1/112291/ivc_full_preprint.pdf?sequence=1&amp;isAllowed=y">Ten-month-old
infants infer the value of goals from the costs of actions</a> -
<strong><em>Science</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=11862940312128630925&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A piece of evidence for children’s capability on
ToM.</p></li>
<li><p><a href="https://www.pnas.org/content/116/36/17747">Origins of
the concepts cause, cost, and goal in prereaching infants</a> -
<strong><em>Proceedings of the National Academy of
Sciences</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=15973074852436355789&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://static1.squarespace.com/static/595a9f155016e1f7ead6edf1/t/61eeb3e7bbc41a23cd288f8a/1643033708945/Gandhi_etal_2021.pdf">Baby
Intuitions Benchmark (BIB): Discerning the goals, preferences, and
actions of others</a> - <strong><em>NeurIPS’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=16514364601966350574">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/2011.05558.pdf">Intentonomy: a
Dataset and Study towards Human Intent Understanding</a> -
<strong><em>CVPR’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=5268870345003195142&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A large-scale database on human intentionally-posted
images on social media.</p></li>
<li><p><a
href="https://www.tshu.io/HeiderSimmel/CogSci20/Flatland_CogSci20.pdf">Adventures
in Flatland: Perceiving Social Interactions Under Physical Dynamics</a>
- <strong><em>CogSci’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=1928005249823745390&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://ojs.aaai.org/index.php/AAAI/article/view/16167">PHASE:
PHysically-grounded Abstract Social Events for Machine Social
Perception</a> - <strong><em>AAAI’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=15536873427310696150&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a href="https://tshu.io/PHASE/">Project</a>].</p></li>
<li><p><a
href="https://openreview.net/forum?id=w_7JMpGZRh0">Watch-And-Help: A
Challenge for Social Perception and Human-AI Collaboration</a> -
<strong><em>ICLR’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=16340001407726295133">All
Versions</a>].</p></li>
<li><p><a href="https://escholarship.org/uc/item/2j53v5nv">Evaluating
and Modeling Social Intelligence: A Comparative Study of Human and AI
Capabilities</a> - <strong><em>CogSci’24</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=902767361177896884">All
Versions</a>]. This work eveloped a comprehensive theoretical framework
for social dynamics and introduced two evaluation tasks: Inverse
Reasoning (IR) and Inverse Inverse Planning (IIP). The approach also
encompassed a computational model based on recursive Bayesian inference,
adept at elucidating diverse human behavioral patterns. Extensive
experiments and detailed analyses revealed that humans surpassed the
latest GPT models in overall performance, zero-shot learning, one-shot
generalization, and adaptability to multi-modalities.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="analogy">Analogy</h3>
<ul>
<li><p><a
href="https://plato.stanford.edu/entries/metaphor/">Metaphor</a> -
<strong><em>Plato Stanford</em></strong>. A computational philosophy
account on Metaphor, a poetically or rhetorically ambitious use of
words, a figurative as opposed to literal use.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/reasoning-analogy/">Analogy and
Analogical Reasoning</a> - <strong><em>Plato Stanford</em></strong>. A
computational philosophy account on Analogy, a comparison between two
objects, or systems of objects, that highlights respects in which they
are thought to be similar.</p></li>
<li><p><a href="https://1lib.net/book/1165963/e9aa3d">A Cognitive Theory
of Metaphor</a> - <strong><em>MIT Press</em></strong>, 1985. [<a
href="https://scholar.google.com/scholar?hl=en&amp;as_sdt=0%2C5&amp;q=a+cognitive+theory+of+metaphor&amp;btnG=">All
Versions</a>]. A cognitive account on Metaphor.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/abs/pii/0004370289900775">The
structure-mapping engine: Algorithm and examples</a> -
<strong><em>Artificial Intelligence</em></strong>, 1989. [<a
href="https://scholar.google.com/scholar?cluster=16104901325436513899&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A computational implementation of analogy.</p></li>
<li><p><a
href="https://cogsci.ucsd.edu/~coulson/203/gentner-markman-97.pdf">Structure
mapping in analogy and similarity</a> - <strong><em>American
Psychologist</em></strong>, 1997. [<a
href="https://scholar.google.com/scholar?cluster=3497411606978611830&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A perspective unifying analogy and similarity
judgement.</p></li>
<li><p><a href="https://psycnet.apa.org/record/2022-26663-001">A theory
of relation learning and cross-domain generalization</a> -
<strong><em>Psychological Review</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=8559821723107269122&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A comprehensive review on the perspective of treating
analogy as cross-domain generalization.</p></li>
<li><p><a
href="https://www.pnas.org/content/pnas/116/10/4176.full.pdf">Emergence
of analogy from relation learning</a> - <strong><em>Proceedings of the
National Academy of Sciences</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=4877125748339538047&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Analogy feature in language models.</p></li>
<li><p><a
href="https://proceedings.mlr.press/v97/allen19a.html">Analogies
Explained: Towards Understanding Word Embeddings</a> -
<strong><em>ICML’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=15445529659618849253&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Explaining the analogy capability in word
embeddings.</p></li>
<li><p><a href="https://aclanthology.org/P17-1007/">Skip-Gram − Zipf +
Uniform = Vector Additivity</a> - <strong><em>ACL’17</em></strong>,
2017. [<a
href="https://scholar.google.com/scholar?cluster=11732363456979525246&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.iiia.csic.es/~enric/papers/generalize_and_blend.pdf">Generalize
and Blend: Concept Blending Based on Generalization, Analogy, and
Amalgams</a> - <strong><em>ICCC’15</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=11073359237116879862&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://proceedings.mlr.press/v28/juhwang13.pdf">Analogy-preserving
Semantic Embedding for Visual Object Categorization</a> -
<strong><em>ICML’13</em></strong>, 2013. [<a
href="https://scholar.google.com/scholar?cluster=9332855910734484101&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The first application of analogy to machine
learning.</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2015/file/45f31d16b1058d586fc3be7207b58053-Paper.pdf">VISALOGY:
Answering Visual Analogy Questions</a> -
<strong><em>NeurIPS’15</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=7665427758655324654&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://ieeexplore.ieee.org/document/9010418">Detecting
Unseen Visual Relations Using Analogies</a> -
<strong><em>CVPR’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=16686853801653819556&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0004370218301863">Analogy
between concepts</a> - <strong><em>Artificial
Intelligence</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=1397905953174123757&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A mathematical account on analogy.</p></li>
<li><p><a href="https://arxiv.org/abs/1902.00120">Learning to Make
Analogies by Contrasting Abstract Relational Structure</a> -
<strong><em>ICLR’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=15521573039503233138&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://aclanthology.org/2020.figlang-1.pdf#page=140">Sky + Fire =
Sunset. Exploring Parallels between Visually Grounded Metaphors and
Image Classifiers</a> - <strong><em>ACL’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=5747285277687442001&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/2006.04156.pdf">Analogy as
Nonparametric Bayesian Inference over Relational Systems</a> -
<strong><em>CogSci’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=1798148167130120057&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.cs.jhu.edu/~alanlab/Pubs21/ichien2021visual.pdf">Visual
Analogy: Deep Learning Versus Compositional Models</a> -
<strong><em>CogSci’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=1187822306970312749&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A human-deep-learning comparison on similarity
judgement.</p></li>
<li><p><a
href="https://escholarship.org/content/qt3j2576vv/qt3j2576vv.pdf">Preschoolers
and adults make inferences from novel metaphors</a> -
<strong><em>CogSci’22</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=16038983545360341739&amp;hl=en&amp;as_sdt=0,44">All
Versions</a>]. A piece of evidence that understanding metaphors is
capable for different cognitive development phases.</p></li>
<li><p><a
href="https://pcl.sitehost.iu.edu/rgoldsto/pdfs/simdiff.pdf">Similarity
involving attributes and relations: Judgments of similarity and
difference are not inverses</a> - <strong><em>Psychological
Science</em></strong>, 1990. [<a
href="https://scholar.google.com/scholar?cluster=13205938250772079784&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="causality">Causality</h3>
<ul>
<li><p><a href="https://en.wikipedia.org/wiki/Causality">Causality</a> -
<strong><em>Wikipedia</em></strong>. Wikipedia on causality, which is
influence by which one event, process, state, or object (a cause)
contributes to the production of another event, process, state, or
object (an effect) where the cause is partly responsible for the effect,
and the effect is partly dependent on the cause.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/causal-models/">Causal
Models</a> - <strong><em>Plato Stanford</em></strong>. A computational
philosophy account on Causal models, which are mathematical models
representing causal relationships within an individual system or
population.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/content-causal/">Causal
Theories of Mental Content</a> - <strong><em>Plato
Stanford</em></strong>. A computational philosophy account on causal
theories of mental content, which attempts to explain how thoughts can
be about things.</p></li>
<li><p><a
href="http://www.jakebowers.org/ITVExperiments/angristimbensrubin96.pdf">Identification
of Causal Effects Using Instrumental Variables</a> - <strong><em>Journal
of the American Statistical Association</em></strong>, 1996. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=17166265099721941605">All
Versions</a>].</p></li>
<li><p><a
href="https://www.psych.uni-goettingen.de/de/cognition/publikationen-dateien-waldmann/1992_predictive_vs_diagnostic.pdf">Predictive
and Diagnostic Learning Within Causal Models: Asymmetries in Cue
Competition</a> - <strong><em>Journal of Experimental
Psychology</em></strong>, 1992. [<a
href="https://scholar.google.com/scholar?cluster=9614241045842043939&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Experimental evidences for distincting causality and
association.</p></li>
<li><p><a
href="https://www.oxfordhandbooks.com/view/10.1093/oxfordhb/9780195376746.001.0001/oxfordhb-9780195376746-e-46">Causal
Reasoning</a> - <strong><em>The Oxford Handbook of Cognitive
Psychology</em></strong>, 2013. [<a
href="https://scholar.google.com/scholar?cluster=11361740093816709089&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://ftp.cs.ucla.edu/pub/stat_ser/R265.pdf">Reasoning
with cause and effect</a> - 1998. Judea Pearl’s tutorials on causal
reasoning with operations on Bayesian networks.</p></li>
<li><p><a href="https://dl.acm.org/doi/pdf/10.1145/3241036">The Seven
Tools of Causal Inference, with Reflections on Machine Learning</a> -
<strong><em>Communications of the ACM</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=13296019510897277617&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Judea Pearl’s review on causal inference in probabilistic
graph models.</p></li>
<li><p><a
href="https://cardiacmr.hms.harvard.edu/files/cardiacmr/files/toward_causal_representation_learning.pdf">Toward
Causal Representation Learning</a> - <strong><em>Proceedings of the
IEEE</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=15629454810797806102&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Yoshua Bengio’s review on the perspective of treating
causal inference as a representation learning problem.</p></li>
<li><p><a
href="https://cocosci.princeton.edu/tom/papers/tbci.pdf">Theory-Based
Causal Induction</a> - <strong><em>Psychological Review</em></strong>,
2009. [<a
href="https://scholar.google.com/scholar?cluster=13980129728092173387&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Thomas Griffiths’ review on causal Bayesian theory
induction.</p></li>
<li><p><a
href="https://ojs.aaai.org//index.php/AAAI/article/view/5483">Theory-Based
Causal Transfer: Integrating Instance-Level Induction and Abstract-Level
Structure Learning</a> - <strong><em>AAAI’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=9411622427165139667&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A computatinoal account on causal transfer.</p></li>
<li><p><a
href="https://onlinelibrary.wiley.com/doi/abs/10.1207/s15516709cog2703_6">Inferring
causal networks from observations and interventions</a> -
<strong><em>Cognitive Science</em></strong>, 2010. [<a
href="https://scholar.google.com/scholar?cluster=12050301037347772984&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://cogsci.mindmodeling.org/2015/papers/0418/paper0418.pdf">Constraints
on Hypothesis Selection in Causal Learning</a> -
<strong><em>CogSci’15</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?hl=en&amp;as_sdt=2005&amp;sciodt=0%2C5&amp;cites=16920774374067505248&amp;scipsc=&amp;q=Constraints+on+hypothesis+selection+in+causal+learning&amp;btnG=">All
Versions</a>].</p></li>
<li><p><a
href="http://cocolab.stanford.edu/papers/GerstenbergEtAl17_PsychScience.pdf">Eye-tracking
causality</a> - <strong><em>Psychological Science</em></strong>, 2017.
[<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=17518200401109470519">All
Versions</a>].</p></li>
<li><p><a
href="https://scholar.google.com/citations?view_op=view_citation&amp;hl=en&amp;user=d0TfP8EAAAAJ&amp;sortby=pubdate&amp;citation_for_view=d0TfP8EAAAAJ:S16KYo8Pm5AC">What
happened? Reconstructing the past through vision and sound</a> - 2021.
[<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=12975579257004398798">All
Versions</a>].</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007/s42113-021-00124-z">How
do people generalize causal relations over objects? A non-parametric
Bayesian account</a> - <strong><em>Computational Brain &amp;
Behavior</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=3364672295201228487">All
Versions</a>]. [<a href="https://psyarxiv.com/x57hf/">Preprint</a>]. How
do people decide how general a causal relationship is, in terms of the
entities or situations it applies to? What features do people use to
decide whether a new situation is governed by a new causal law or an old
one? How can people make these difficult judgments in a fast, efficient
way? This paper addresses these questions in two experiments that ask
participants to generalize from one (Experiment 1) or several
(Experiment 2) causal interactions between pairs of objects. In each
case, participants see an agent object act on a recipient object,
causing some changes to the recipient.</p></li>
<li><p><a
href="https://www.psych.uni-goettingen.de/de/cognition/publikationen-dateien-waldmann/2006_science.pdf">Causal
Reasoning in Rats</a> - <strong><em>Science</em></strong>, 2006. [<a
href="https://scholar.google.com/scholar?cluster=17987039255457850949&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A piece of evidence for the capability of causal
reasoning in intelligent animals.</p></li>
<li><p><a
href="http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.183.4674&amp;rep=rep1&amp;type=pdf">Do
New Caledonian crows solve physical problems through causal
reasoning?</a> - <strong><em>Proceedings of the Royal Society B:
Biological Sciences</em></strong>, 2009. [<a
href="https://scholar.google.com/scholar?cluster=18374985546068164189&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A piece of evidence for the capability of causal
reasoning in intelligent animals.</p></li>
<li><p><a href="http://fitelson.org/woodward/leslie.pdf">Do
six-month-old infants perceive causality?</a> -
<strong><em>Cognition</em></strong>, 1987. [<a
href="https://scholar.google.com/scholar?cluster=14270905342434182186&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="commonsense">Commonsense</h3>
<h4 id="intuitive-physics">Intuitive Physics</h4>
<ul>
<li><p><a
href="https://github.com/lishiqianhugh/Intuitive_Physics_Reading_List">Intuitive
Physics Reading List</a> - <strong><em>GitHub</em></strong>. A reading
list on intuitive physics, maintained actively by Shiqian Li.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S1364661317301262">Intuitive
Physics: Current Research and Controversies</a> - <strong><em>Trends in
Cognitive Sciences</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?start=0&amp;hl=en&amp;as_sdt=0,5&amp;cluster=12085981794958916203">All
Versions</a>]. Hongjing Lu’s review on intuitive physics.</p></li>
<li><p><a
href="https://www.pnas.org/content/pnas/110/45/18327.full.pdf">Simulation
as an engine of physical scene understanding</a> -
<strong><em>Proceedings of the National Academy of
Sciences</em></strong>, 2013. [<a
href="https://scholar.google.com/scholar?cluster=5892822406285231676&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://www.pnas.org/content/pnas/suppl/2013/10/18/1306572110.DCSupplemental/pnas.201306572SI.pdf?targetid=nameddest%3DSTXT">Appendix</a>].
The first attempt to computationally simulate intuitive
physics.</p></li>
<li><p><a
href="https://www.pnas.org/doi/pdf/10.1073/pnas.1610344113">Functional
neuroanatomy of intuitive physical inference</a> -
<strong><em>Proceedings of the National Academy of
Sciences</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=1792195093536891402&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A piece of evidence for the functional part of intuitive
physics in human brain.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S1364661317301134">Mind
Games: Game Engines as an Architecture for Intuitive Physics</a> -
<strong><em>Trends in Cognitive Sciences</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=14527964477161848029&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>]. Tomer Ullman’s review on simulation-based intuitive
physics.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/abs/pii/S0010028517301822">Learning
physical parameters from dynamic scenes</a> - <strong><em>Cognitive
Psychology</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=5103729321433959736&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0010028521000190">Limits
on Simulation Approaches in Intuitive Physics</a> -
<strong><em>Cognitive Psychology</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=6329029167380621767">All
Versions</a>]. Ernest Davis’s perspective against intuitive physics,
that physcial reasoning is logical reasoning instead of
intuition.</p></li>
<li><p><a href="https://psyarxiv.com/y4a8x/download?format=pdf">Partial
Mental Simulation Explains Fallacies in Physical Reasoning</a> -
<strong><em>Cognitive Neuropsychology</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=15541954459060383152&amp;hl=en&amp;as_sdt=2005">All
Versions</a>].</p></li>
<li><p><a
href="https://www.nature.com/articles/s41562-022-01394-8">Intuitive
physics learning in a deep-learning model inspired by developmental
psychology</a> - <strong><em>Nature Human Behavior</em></strong>, 2022.
[<a
href="https://scholar.google.com/scholar?cluster=13803979681049451699&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A machine-learning dataset designed to evaluate
conceptual understanding of intuitive physics, adopting the
violation-of-expectation (VoE) paradigm from developmental psychology; a
deep-learning system that learns intuitive physics directly from visual
data, inspired by studies of visual cognition in children.</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2019/hash/4191ef5f6c1576762869ac49281130c9-Abstract.html">PHYRE:
A New Benchmark for Physical Reasoning</a> -
<strong><em>NeurIPS’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=9555658528231205655&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A benchmark for AI physical reasoning.</p></li>
<li><p><a
href="https://www.nature.com/articles/s42256-022-00583-4">Phy-Q as a
measure for physical reasoning intelligence</a> - <strong><em>Nature
Machine Intelligence</em></strong>, 2023. [<a
href="https://www.nature.com/articles/s42256-019-0072-x">NMI
Challenge</a>]. An interactive benchmark for AI physical
reasoning.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="ai-commonsense-reasoning">AI Commonsense Reasoning</h4>
<ul>
<li><p><a
href="https://www.sciencedirect.com/book/9781483207704/representations-of-commonsense-knowledge">Representations
of Commonsense Knowledge</a> - <strong><em>Morgan
Kaufmann</em></strong>, 1990. [<a
href="https://scholar.google.com/scholar?cluster=8861902735724600978">All
Versions</a>]. A classic book on commonsense knowledge.</p></li>
<li><p><a
href="https://link.springer.com/chapter/10.1007%2F3-540-53487-3_59">Towards
a theory of commonsense visual reasoning</a> -
<strong><em>FSTTCS</em></strong>, 1990. [<a
href="https://scholar.google.com/scholar?cluster=13178231862265713961&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper on visual commonsense.</p></li>
<li><p><a
href="http://cs.wellesley.edu/~cs125/reading/commonsenseAI.pdf">Commonsense
reasoning and commonsense knowledge in artificial intelligence</a> -
<strong><em>Communications of the ACM</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=13786590180441485203&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Gary Marcus’s review on commonsense knowledge in
AI.</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&amp;arnumber=8953217">From
Recognition to Cognition: Visual Commonsense Reasoning</a> -
<strong><em>CVPR’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=15467433880059136365&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="http://visualcommonsense.com/">Project</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/1911.11641.pdf">PIQA: Reasoning
about Physical Commonsense in Natural Language</a> -
<strong><em>AAAI’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=10110424163152713144&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&amp;arnumber=9156347">Visual
Commonsense R-CNN</a> - <strong><em>CVPR’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=6886229776034162585&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://openreview.net/pdf?id=Byg1v1HKDB">Abductive
Commonsense Reasoning</a> - <strong><em>ICLR’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=16544200144479839958&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Abductive commonsense reasoning on large language
models.</p></li>
<li><p><a
href="https://link.springer.com/chapter/10.1007%2F978-3-030-58558-7_30">VisualCOMET:
Reasoning About the Dynamic Context of a Still Image</a> -
<strong><em>ECCV’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=7681600847940772451&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://link.springer.com/chapter/10.1007/978-3-031-20059-5_32">The
Abduction of Sherlock Holmes: A Dataset for Visual Abductive
Reasoning</a> - <strong><em>ECCV’22</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=18355807581692234364">All
Versions</a>]. [<a
href="https://arxiv.org/abs/2202.04800">Preprint</a>]. This paper
presents Sherlock, an annotated corpus of 103K images for testing
machine capacity for abductive reasoning beyond literal image contents.
The corpus construction process adopts a free-viewing paradigm:
participants first observe and identify salient clues within images
(e.g., objects, actions) and then provide a plausible inference about
the scene, given the clue.</p></li>
<li><p><a
href="https://aclanthology.org/2024.naacl-long.469/">UNcommonsense
Reasoning: Abductive Reasoning about Uncommon Situations</a> -
<strong><em>NAACL’24</em></strong>, 2024. [<a
href="https://scholar.google.com/scholar?cluster=470445696014235795">All
Versions</a>]. This paper explores the task of uncommonsense abductive
reasoning. Given a piece of context with an unexpected outcome, this
task requires reasoning abductively to generate an explanation that
makes the unexpected outcome more likely in the context.</p></li>
<li><p><a
href="https://aclanthology.org/2020.emnlp-main.703.pdf">Experience
Grounds Language</a> - <strong><em>EMNLP’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=3734668471751920487&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A perspective on the furture of computational linguistics
research—commonsense-driven and embodied language.</p></li>
<li><p><a href="https://aclanthology.org/2021.emnlp-main.162/">Broaden
the Vision: Geo-Diverse Visual Commonsense Reasoning</a> -
<strong><em>EMNLP’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=12305856131717604775&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://www.charleskemp.com/papers/hanrpk_humanlikepropertyinductionisachallengeforlargelanguagemodels.pdf">Human-like
property induction is a challenge for large language models</a> -
<strong><em>CogSci’22</em></strong>, 2022.</p></li>
<li><p><a href="https://arxiv.org/abs/2305.17390">SwiftSage: A
Generative Agent with Fast and Slow Thinking for Complex Interactive
Tasks</a> - <strong><em>NeurIPS’23</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=3844178012869500706&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://swiftsage.github.io/">Project</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h4 id="commonsense-knowledgebase">Commonsense Knowledgebase</h4>
<ul>
<li><p><a href="https://www.wikihow.com/Main-Page">wikiHow</a> -
<strong><em>wikiHow.com</em></strong>. wikiHow is on website hosting
step-by-step “How-to” procedural instructions across various domains and
topics.</p></li>
<li><p><a href="https://theworldavatar.io/">The World Avatar</a> -
<strong><em>The World Avatar™</em></strong>. A large-scale dynamic
knowledge graph connecting concepts with relations to digitalize
molecules, buildings, cities, and countries.</p></li>
<li><p><a
href="https://faculty.cc.gatech.edu/~isbell/classes/reading/papers/lenat95cyc.pdf">CYC:
A Large-Scale Investment in Knowledge Infrastructure</a> -
<strong><em>Communications of the ACM</em></strong>, 1995. [<a
href="https://scholar.google.com/scholar?cluster=6505009388871605141&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The first attempt to build large-scale commonse
knoweldgebase from human knowledge.</p></li>
<li><p><a href="https://arxiv.org/pdf/1612.03975.pdf">ConceptNet 5.5: An
Open Multilingual Graph of General Knowledge</a> -
<strong><em>AAAI’17</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=7089916805257737701&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Latest version of ConceptNet.</p></li>
<li><p><a
href="https://www.aaai.org/Library/Symposia/Spring/2002/ss02-09-011.php">The
Public Acquisition of Commonsense Knowledge</a> -
<strong><em>Proceedings of AAAI Spring Symposium on Acquiring (and
Using) Linguistic (and World) Knowledge for Information
Access</em></strong>, 2002. [<a
href="https://scholar.google.com/scholar?cluster=12533779219524472080&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The first attempt for acquring commonsense knowlege from
humans’ activities on the internet.</p></li>
<li><p><a
href="http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.472.914&amp;rep=rep1&amp;type=pdf">Open
Mind Common Sense: Knowledge Acquisition from the General Public</a> -
<strong><em>OTM Confederated International Conferences’02</em></strong>,
2002. [<a
href="https://scholar.google.com/scholar?cluster=11431785236825227404&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]..</p></li>
<li><p><a
href="http://www.aladdin.cs.cmu.edu/papers/pdfs/y2006/verbosity.pdf">Verbosity:
A Game for Collecting Common-Sense Facts</a> -
<strong><em>CHI’06</em></strong>, 2006. [<a
href="https://scholar.google.com/scholar?cluster=7793704394155465847&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://dl.acm.org/doi/fullHtml/10.1145/1378704.1378719">Designing
games with a purpose</a> - <strong><em>Communications of the
ACM</em></strong>, 2008. [<a
href="https://scholar.google.com/scholar?cluster=18332117920150730595&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://people.mpi-inf.mpg.de/~ntandon/papers/aaai-2014-tandon.pdf">Acquiring
Comparative Commonsense Knowledge from the Web</a> -
<strong><em>AAAI’14</em></strong>, 2014. [<a
href="https://scholar.google.com/scholar?cluster=16641273554706459553&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/abstract/document/9904017">Visual
Concept Programming: A Visual Analytics Approach to Injecting Human
Intelligence at Scale</a> - <strong><em>IEEE Transactions on
Visualization and Computer Graphics</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=10724509334112758172&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. This paper presents Visual Concept Programming, a
first-of-its-kind visual analytics approach of using visual concepts to
program image data at scale while requiring a few human
efforts.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="inductive-logic-program-synthesis">Inductive Logic &amp; Program
Synthesis</h3>
<ul>
<li><p><a
href="https://plato.stanford.edu/entries/logic-inductive/">Inductive
Logic</a> - <strong><em>Plato Stanford</em></strong>. A computational
philosophy account on Inductive Logic, which is a logic of evidential
support.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/modeltheory-fo/">First-order
Model Theory</a> - <strong><em>Plato Stanford</em></strong>. A
computational philosophy account on First-order Model Theory, which is a
branch of mathematics that deals with the relationships between
descriptions in first-order languages and the structures that satisfy
these descriptions.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/logic-paraconsistent/">Paraconsistent
Logic</a> - <strong><em>Plato Stanford</em></strong>. A computational
philosophy account on Paraconsistent Logic, where any logic is
paraconsistent as long as it is not explosive.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/logical-consequence/">Logical
Consequence</a> - <strong><em>Plato Stanford</em></strong>. A
computational philosophy account on Logical Consequence, which is about
the relation between premises and conclusions in valid
arguments.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/logical-pluralism/">Logic
Pluralism</a> - <strong><em>Plato Stanford</em></strong>. A
computational philosophy account on Logic Pluralism, which is the view
that there is more than one correct logic.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/logic-firstorder-emergence/">The
Emergence of First-Order Logic</a> - <strong><em>Plato
Stanford</em></strong>. A computational philosophy account on the
emergence of first-order logic, mainly about first-order logic is
natural retrospect.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/logic-higher-order/">Second-order
and Higher-order Logic</a> - <strong><em>Plato
Stanford</em></strong>.</p></li>
<li><p><a
href="https://www.microsoft.com/en-us/research/wp-content/uploads/2017/10/program_synthesis_now.pdf">Program
Synthesis</a> - <strong><em>Foundations and Trends in Programming
Languages</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=5442933587668978421&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Sumit Gulwani’s comprehensive review on program
synthesis.</p></li>
<li><p><a
href="https://www.ijcai.org/Proceedings/83-1/Papers/109.pdf">The
Discovery of the Equator or Concept Driven Learning</a> -
<strong><em>IJCAI’83</em></strong>, 1983. [<a
href="https://scholar.google.com/scholar?cluster=15712225225140903169&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper on second-order metarules.</p></li>
<li><p><a
href="https://link.springer.com/chapter/10.1007%2F3-540-44797-0_10">Towards
combining inductive logic programming with Bayesian networks</a> -
<strong><em>ILP’01</em></strong>, 2001. [<a
href="https://scholar.google.com/scholar?cluster=2904180673047700407&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://www.doc.ic.ac.uk/~shm/Papers/metagol_gram.pdf">Meta-interpretive
learning: application to grammatical inference</a> - <strong><em>Machine
Learning</em></strong>, 2014. [<a
href="https://scholar.google.com/scholar?cluster=17075313112718885592&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Stephen Muggleton’s original paper on Meta-Interpretive
Learning (MIL).</p></li>
<li><p><a
href="http://andrewcropper.com/pubs/ijcai15-metagolo.pdf">Learning
Efficient Logical Robot Strategies Involving Composable Objects</a> -
<strong><em>IJCAI’15</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=5109851972354087162&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://andrewcropper.com/pubs/ijcai16-metafunc.pdf">Learning
Higher-Order Logic Programs through Abstraction and Invention</a> -
<strong><em>IJCAI’16</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=10945054943203858325&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://link.springer.com/chapter/10.1007%2F978-3-319-99960-9_3">How
Much Can Experimental Cost Be Reduced in Active Learning of Agent
Strategies?</a> - <strong><em>ILP’18</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=8152380236842970357&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007/s10994-018-5710-8">Meta-Interpretive
Learning from noisy images</a> - <strong><em>Machine
Learning</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=5719375383968868329&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://andrewcropper.com/pubs/mlj18-metaopt.pdf">Learning
efficient logic programs</a> - <strong><em>Machine
Learning</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=17955696870252443734&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="http://andrewcropper.com/pubs/mlj19-metaho.pdf">Learning
higher-order logic programs</a> - <strong><em>Machine
Learning</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=6723896359456002413&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="http://andrewcropper.com/pubs/mlj19-reduce.pdf">Logical
reduction of metarules</a> - <strong><em>Machine Learning</em></strong>,
2019. [<a
href="https://scholar.google.com/scholar?cluster=4577603126537024540&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://andrewcropper.com/pubs/ijcai19-playgol.pdf">Playgol:
Learning Programs Through Play</a> - <strong><em>IJCAI’19</em></strong>,
2019. [<a
href="https://scholar.google.com/scholar?cluster=556522464212000763&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://link.springer.com/article/10.1007%2Fs00354-019-00054-2">Machine
Discovery of Comprehensible Strategies for Simple Games Using
Meta-interpretive Learning</a> - <strong><em>New Generation
Computing</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=11019349634035542991&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://andrewcropper.com/pubs/aaai20-forgetgol.pdf">Forgetting to
Learn Logic Programs</a> - <strong><em>AAAI’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=13676986733133377042&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://www.ijcai.org/proceedings/2020/673">Turning 30:
New Ideas in Inductive Logic Programming</a> -
<strong><em>IJCAI’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=17980870844719684257&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/2008.07912">Inductive logic
programming at 30: a new introduction</a> - <strong><em>Journal of
Artificial Intelligence Research</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=317114056670544302&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A 30-year comprehensive review on Inductive Logic
Programming.</p></li>
<li><p><a href="https://arxiv.org/pdf/2005.02259.pdf">Learning programs
by learning from failures</a> - <strong><em>Machine
Learning</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=6797200487935462023&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://www.ijcai.org/proceedings/2020/320">Complete
Bottom-Up Predicate Invention in Meta-Interpretive Learning</a> -
<strong><em>IJCAI’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=6085183078630665234&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/2106.07464.pdf">Meta-Interpretive
Learning as Metarule Specialisation</a> - <strong><em>Machine
Learning</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=14684315775211086859&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0004370204000591">Qualitative
choice logic</a> - <strong><em>Artificial Intelligence</em></strong>,
2004. [<a
href="https://scholar.google.com/scholar?cluster=1586187056162326386&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.ijcai.org/Proceedings/16/Papers/278.pdf">Derivative-free
optimization of high-dimensional non-convex functions by sequential
random embeddings</a> - <strong><em>IJCAI’16</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=15955040483290586781&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://londmathsoc.onlinelibrary.wiley.com/doi/abs/10.1112/S0024610704006106">Finitely
Generated Groups and First-Order Logic</a> - <strong><em>Journal of The
London Mathematical Society-second Series</em></strong>, 2005. [<a
href="https://scholar.google.com/scholar?cluster=3457158221419711506&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://vigilworkshop.github.io/static/papers-2021/25.pdf">Leveraging
Language for Abstraction and Program Search</a> -
<strong><em>ICML’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?hl=en&amp;as_sdt=0%2C5&amp;q=Leveraging+Language+for+Abstraction+and+Program+Search&amp;btnG=">All
Versions</a>].</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2021/hash/f7e2b2b75b04175610e5a00c1e221ebb-Abstract.html">Program
Synthesis Guided Reinforcement Learning</a> -
<strong><em>NeurIPS’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=17353674428642875269&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://cogtoolslab.github.io/pdf/wang_cogsci_2021a.pdf">Learning
Part-Based Abstractions for Visual Object Concepts</a> -
<strong><em>CogSci’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?lookup=0&amp;q=Learning+Part-Based+Abstractions+for+Visual+Object+Concepts&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/2108.07732">Program Synthesis with
Large Language Models</a> - 2021. [<a
href="https://scholar.google.com/scholar?cluster=15213050540818392833">All
Versions</a>]. This paper explores the limits of the current generation
of large language models for program synthesis in general purpose
programming languages.</p></li>
<li><p><a href="https://dl.acm.org/doi/abs/10.1145/3571249">Combining
Functional and Automata Synthesis to Discover Causal Reactive
Programs</a> - <strong><em>POPL’23</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=10470162446663474225&amp;as_sdt=0,5">All
Versions</a>]. A new algorithm that synthesizes functional reactive
programs from observation data, which iterates between a functional
synthesis step, which attempts to generate a transition function over
observed states, and an automata synthesis step, which adds any
additional latent state necessary to fully account for the
observations.</p></li>
<li><p><a
href="http://cap.csail.mit.edu/sites/default/files/research-pdfs/Synthesizing%20theories%20of%20human%20language%20with%20Bayesian%20program%20induction.pdf">Synthesizing
theories of human language with Bayesian program induction</a> -
<strong><em>Nature Communications</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?cluster=8603772394100237159&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/2306.12672">From Word Models to
World Models: Translating from Natural Language to the Probabilistic
Language of Thought</a> - 2023. [<a
href="https://scholar.google.com/scholar?cluster=13778788929096574993">All
Versions</a>]. Rational meaning construction, a computational framework
for language-informed thinking that combines neural language models with
probabilistic models for rational inference. Linguistic meaning is
framed as a context-sensitive mapping from natural language into a
probabilistic language of thought (PLoT)–a general-purpose symbolic
substrate for generative world modeling.</p></li>
<li><p><a href="https://proceedings.mlr.press/v139/hong21a.html">Latent
Programmer: Discrete Latent Codes for Program Synthesis</a> -
<strong><em>ICML’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=9789877360194738968">All
Versions</a>]. Paper introducing the Latent Programmer, a two-level
program synthesis method that first predicts a discrete latent code from
input/output examples, and then generates the program in the target
language.</p></li>
<li><p><a href="https://proceedings.mlr.press/v202/gao23f">PAL:
Program-aided Language Models</a> - <strong><em>ICML’23</em></strong>,
2023. [<a
href="https://scholar.google.com/scholar?cluster=14898051625978777315&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Paper presenting an approach that uses the LLM to read
natural language problems and generate programs as the intermediate
reasoning steps, but offloads the solution step to a runtime such as a
Python interpreter. With PAL, decomposing the natural language problem
into runnable steps remains the only learning task for the LLM, while
solving is delegated to the interpreter.</p></li>
<li><p><a href="https://aclanthology.org/2023.acl-long.411/">Large
Language Models Meet NL2Code: A Survey</a> -
<strong><em>ACL’23</em></strong>, 2023. [<a
href="https://scholar.google.com/scholar?cluster=11868015824802341463&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a href="https://nl2code.github.io/">NL2Code
Website</a>]. A paper presenting a comprehensive survey of 27 existing
large language models for NL2Code, and also review benchmarks and
metrics, suggesting that the key factors contributing to the success of
large language models for NL2Code are “Large Size, Premium Data, Expert
Tuning”.</p></li>
<li><p><a href="https://dl.acm.org/doi/abs/10.1145/3597503.3608128">A
Large-Scale Survey on the Usability of AI Programming Assistants:
Successes and Challenges</a> - <strong><em>ICSE’24</em></strong>, 2024.
[<a
href="https://scholar.google.com/scholar?cluster=3696356619002071917&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A survey finding that developers are most motivated to
use AI programming assistants because they help developers reduce
key-strokes, finish programming tasks quickly, and recall syntax, but
resonate less with using them to help brainstorm potential
solutions.</p></li>
<li><p><a href="https://arxiv.org/abs/2308.10620">Large Language Models
for Software Engineering: A Systematic Literature Review</a> - 2023. [<a
href="https://scholar.google.com/scholar?cluster=10466731638053452642&amp;as_sdt=0,5">All
Versions</a>]. A systematic literature review on LLM4SE, with a
particular focus on understanding how LLMs can be exploited to optimize
processes and outcomes.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="knowledge-representation">Knowledge Representation</h3>
<ul>
<li><p><a href="https://1lib.net/book/511192/9eab86">Handbook of
Knowledge Representation</a> - <strong><em>Elsevier</em></strong>, 2008.
[<a
href="https://scholar.google.com/scholar?cluster=14732064619564679879&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A pragmatical handbook for all kinds of knowledge
representation modes.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/logic-ontology/">Logic and
Ontology</a> - <strong><em>Plato Stanford</em></strong>. A computational
philosophy account on logic and ontology, mainly about the intersections
of logic and ontology in many significant philosophy problems.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/language-thought/">The Language
of Thought Hypothesis</a> - <strong><em>Plato Stanford</em></strong>. A
computational philosophy account on the laugnage of though hypothesis,
which proposes that thinking occurs in a mental language.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/knowledge-analysis/">The
Analysis of Knowledge</a> - <strong><em>Plato
Stanford</em></strong>.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/scientific-representation/">Scientific
Representation</a> - <strong><em>Plato Stanford</em></strong>. A
computational philosophy account on scientific representation, focusing
on how scientific models represent their target systems.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/self-knowledge/">Self-Knowledge</a>
- <strong><em>Plato Stanford</em></strong>. A computational philosophy
account on self-knowledge, which standardly refers to knowledge of one’s
own mental states—that is, of what one is feeling or thinking, or what
one believes or desires.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/common-knowledge/">Common
Knowledge</a> - <strong><em>Plato Stanford</em></strong>.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/sense-data/">Sense-Data</a> -
<strong><em>Plato Stanford</em></strong>.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/supervenience/">Supervenience</a>
- <strong><em>Plato Stanford</em></strong>. A computational philosophy
account on supervenience, where a set of properties A supervenes upon
another set B just in case no two things can differ with respect to
A-properties without also differing with respect to their
B-properties.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/logic-dialogical/">Dialogical
Logic</a> - <strong><em>Plato Stanford</em></strong>. A computational
philosophy account on dialogical logic, which is a dialogue-based
approach to logic and argumentation rooted in a research tradition that
goes back to dialectics in Greek Antiquity, when problems were
approached through dialogues in which opposing parties discussed a
thesis through questions and answers.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/logic-temporal/">Temporal
Logic</a> - <strong><em>Plato Stanford</em></strong>.</p></li>
<li><p><a href="https://plato.stanford.edu/entries/logic-modal/">Modal
Logic</a> - <strong><em>Plato Stanford</em></strong>. A computational
philosophy account on Modal Logic, which is the study of the deductive
behavior of the expressions ‘it is necessary that’ and ‘it is possible
that’.</p></li>
<li><p><a
href="https://plato.stanford.edu/entries/logic-epistemic/">Epistemic
Logic</a> - <strong><em>Plato Stanford</em></strong>. A computational
philosophy account on Epistemic Logic, which is a subfield of
epistemology concerned with logical approaches to knowledge, belief and
related notions.</p></li>
<li><p><a
href="https://en.wikipedia.org/wiki/Epistemic_modal_logic">Epistemic
Modal Logic</a> - <strong><em>Wikipedia</em></strong>.</p></li>
<li><p><a
href="https://perception.jhu.edu/files/PDFs/21_Relations/HafriFirestone_2021_SeeingRelations_TiCS.pdf">The
Perception of Relations</a> - <strong><em>Trends in Cognitive
Sciences</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=12190078466818849725&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>]. Chaz Firestone’s review on the perception of relation, in
constrast to the conventional reasoning view.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/abs/pii/0004370284900390">Commonsense
reasoning about causality: Deriving behavior from structure</a> -
<strong><em>Artificial Intelligence</em></strong>, 1984. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=14940738362673077704">All
Versions</a>].</p></li>
<li><p><a
href="https://link.springer.com/article/10.1023/B:SYNT.0000024912.56773.5e">Logics
for Epistemic Programs</a> - <strong><em>Synthese</em></strong>, 2004.
[<a
href="https://scholar.google.com/scholar?cluster=11403619699670839488&amp;hl=en&amp;as_sdt=0,5&amp;as_vis=1">All
Versions</a>].</p></li>
<li><p><a href="https://tomgruber.org/writing/ontolingua-kaj-1993.pdf">A
Translation Approach to Portable Ontology Specifications</a> -
<strong><em>Knowledge Acquisition</em></strong>, 1993. [<a
href="https://scholar.google.com/scholar?cluster=14668658395073605123&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://www.cs.ox.ac.uk/activities/ieg/e-library/sources/harnad90_sgproblem.pdf">The
Symbolic Grounding Problem</a> - <strong><em>Physica D: Nonlinear
Phenomena</em></strong>, 1990. [<a
href="https://scholar.google.com/scholar?cluster=6279614024681929496&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1467-7687.2007.00585.x?__cf_chl_captcha_tk__=pmd_Q6xVT1AstoEUxA7xS3_10HyDVsk8W_DzWgOPho_Njnw-1635210931-0-gqNtZGzNA1CjcnBszQvl">Learning
overhypotheses with hierarchical Bayesian models</a> -
<strong><em>Developmental Science</em></strong>, 2007. [<a
href="https://scholar.google.com/scholar?cluster=18041836774924845900&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://escholarship.org/content/qt19v2r2ws/qt19v2r2ws.pdf">Learning
Causal Schemata</a> - <strong><em>CogSci’07</em></strong>, 2007, [<a
href="https://scholar.google.com/scholar?cluster=5008191267417189643&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://www.pnas.org/content/105/31/10687">The discovery
of structural form</a> - <strong><em>Proceedings of the National Academy
of Sciences</em></strong>, 2008. [<a
href="https://scholar.google.com/scholar?cluster=10433149156915110486&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Chales Kemp’s review on theory induction.</p></li>
<li><p><a
href="https://onlinelibrary.wiley.com/doi/full/10.1080/03640210701802071">A
Rational Analysis of Rule-Based Concept Learning</a> -
<strong><em>Cognitive Science</em></strong>, 2008. [<a
href="https://scholar.google.com/scholar?cluster=7765061503727822620&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://escholarship.org/content/qt50r1c7qh/qt50r1c7qh.pdf">Modeling
semantic cognition as logical dimensionality reduction</a> -
<strong><em>CogSci’08</em></strong>, 2008. [<a
href="https://scholar.google.com/scholar?cluster=17061801746839695691&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="http://www.charleskemp.com/papers/KempGT08.pdf">Theory
Acquisition and the Language of Thought</a> -
<strong><em>CogSci’08</em></strong>, 2008. [<a
href="https://scholar.google.com/scholar?cluster=1839916602381147749&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://web.mit.edu/tomeru/www/papers/tlss2010.pdf">Theory
Acquisition as Stochastic Search</a> -
<strong><em>CogSci’10</em></strong>, 2010. [<a
href="https://scholar.google.com/scholar?cluster=16324634056226561429&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="http://www.charleskemp.com/papers/kemptng09.pdf">A
probabilistic model of theory formation</a> -
<strong><em>Cognition</em></strong>, 2010. [<a
href="https://scholar.google.com/scholar?cluster=7705799129887482041&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://core.ac.uk/display/78064072">Bootstrapping in a
language of thought: A formal model of numerical concept learning</a> -
<strong><em>Cognition</em></strong>, 2012. [<a
href="https://scholar.google.com/scholar?cluster=13046606910781656302&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://cbmm-dev.mit.edu/sites/default/files/publications/CBMM-Memo-010.pdf">Concepts
in a Probabilistic Language of Thought</a> - <strong><em>Center for
Brains, Minds, and Machines MEMO No.010</em></strong>, 2014. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=14593712389828476130">All
Versions</a>].</p></li>
<li><p><a
href="http://www.charleskemp.com/papers/kemp_exploringtheconceptualuniverse.pdf">Exploring
the Conceptual Universe</a> - <strong><em>Psychological
Review</em></strong>, 2012. [<a
href="https://scholar.google.com/scholar?cluster=17824067813343816306&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://www.charleskemp.com/papers/kempj_ataxonomyofinductiveproblems.pdf">A
taxonomy of inductive problems</a> - <strong><em>Psychonomic Bulletin
&amp; Review</em></strong>, 2014. [<a
href="https://scholar.google.com/scholar?cluster=2571009743105592927&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://colala.berkeley.edu/papers/piantadosi2016logical.pdf">The
Logical Primitives of Thought: Empirical Foundations for Compositional
Cognitive Models</a> - <strong><em>Psychological Review</em></strong>,
2016. [<a
href="https://scholar.google.com/scholar?cluster=5316027496661813145&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://onlinelibrary.wiley.com/doi/full/10.1111/cogs.12580">The
Emergence of Organizing Structure in Conceptual Representation</a> -
<strong><em>Cognitive Science</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=4986316323923233074&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://cogtoolslab.github.io/pdf/wang_cogsci_2021b.pdf">Theory
Acquisition as Constraint-Based Program Synthesis</a> -
<strong><em>CogSci’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=525148607069840280&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://escholarship.org/uc/item/9j00x928">Connecting
perceptual and procedural abstractions in physical construction</a> -
<strong><em>CogSci’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?hl=en&amp;as_sdt=0%2C5&amp;q=Connecting+perceptual+and+procedural+abstractions+in+physical+construction&amp;btnG=">All
Versions</a>].</p></li>
<li><p><a
href="https://www.biorxiv.org/content/10.1101/2021.03.19.385641v1.full.pdf">Invariant
representation of physical stability in the human brain</a> - 2021. [<a
href="https://scholar.google.com/scholar?cluster=17431019238600295521&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.146.4086&amp;rep=rep1&amp;type=pdf">Introduction
to The Fluent Calculus</a> - <strong><em>Linkoeping University
Electronic Press</em></strong>, 1998. [<a
href="https://scholar.google.com/scholar?cluster=12069059079023496731&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0004370299000338">From
situation calculus to fluent calculus: State update axioms as a solution
to the inferential frame problem</a> - <strong><em>Artificial
Intelligence</em></strong>, 1999. [<a
href="https://scholar.google.com/scholar?cluster=10854895617698839149&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://www.stat.ucla.edu/~sczhu/papers/Conf_2013/Learning_AoG_NeurIPS_2013.pdf">Unsupervised
Structure Learning of Stochastic And-Or Grammars</a> -
<strong><em>NeurIPS’13</em></strong>, 2013. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=4354984630817844670">All
Versions</a>].</p></li>
<li><p><a href="https://psyarxiv.com/ysndt">Algorithms of Adaptation in
Inductive Inference</a> - <strong><em>Cognitive
Psychology</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=16222039361294164246&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/0010027795006743">A
representational analysis of numeration systems</a> -
<strong><em>Cognition</em></strong>, 1995. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=8852566070856662412">All
Versions</a>].</p></li>
<li><p><a
href="https://openaccess.thecvf.com/content/CVPR2022/html/Papadopoulos_Learning_Program_Representations_for_Food_Images_and_Cooking_Recipes_CVPR_2022_paper.html">Learning
Program Representations for Food Images and Cooking Recipes</a> -
<strong><em>CVPR’22</em></strong>, 2022. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=7690010749576063125">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/2205.07455">Reasoning about
Procedures with Natural Language Processing: A Tutorial</a> - 2023. [<a
href="https://scholar.google.com/scholar?cluster=11364086808527515615&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="cognitive-development">Cognitive Development</h3>
<ul>
<li><p><a href="https://arxiv.org/abs/1810.07528">Machine Common Sense
Concept Paper</a> - <strong><em>DARPA</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=1603121108181262769&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. DARPA’s perspective on integrating core knowledge from
development psychology into machine intelligence systems.</p></li>
<li><p><a
href="https://en.wikipedia.org/wiki/Cognitive_development">Cognitive
Development</a> - <strong><em>Wikipedia</em></strong>.</p></li>
<li><p><a
href="https://scholar.google.com/scholar?hl=en&amp;as_sdt=0%2C5&amp;q=Cognitive+Development%3A+an+information+processing+approach&amp;btnG=">Cognitive
development: An information processing approach</a> -
<strong><em>B.Blackwell</em></strong>, 1991. [<a
href="https://scholar.google.com/scholar?hl=en&amp;as_sdt=0%2C5&amp;q=Cognitive+development%3A+An+information+processing+approach&amp;btnG=">All
Versions</a>].</p></li>
<li><p><a
href="https://psycnet.apa.org/record/2012-12791-001">Reconstructing
constructivism: Causal models, Bayesian learning mechanisms, and the
theory theory</a> - <strong><em>Psychological Bulletin</em></strong>,
2012. [<a
href="https://scholar.google.com/scholar?cluster=11218217347365817167&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Alison Gopnik’s review on the constructivism idea of
developmental research.</p></li>
<li><p><a
href="https://doi.apa.org/doiLanding?doi=10.1037/rev0000153">Towards a
rational constructivist theory of cognitive development</a> -
<strong><em>Psychological Review</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=3294824172745724080&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Fei Xu’s review extending Gopnik’s view of
constructivism, with the rationality as constraint.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S1364661312001301">The
origins of inquiry: inductive inference and exploration in early
childhood</a> - <strong><em>Trends in Cognitive Sciences</em></strong>,
2012. [<a
href="https://scholar.google.com/scholar?cluster=5189329081728071335&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Laura Schulz’s review on children’s exploratory
play.</p></li>
<li><p><a
href="https://www.annualreviews.org/doi/abs/10.1146/annurev-devpsych-070120-014806">Play,
Curiosity, and Cognition</a> - <strong><em>Annual Review of
Developmental Psychology</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=10278208468154249192&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>]. Laura Schulz’s review on children’s exploratory play,
which proposes a new perspective on exploratory play to explain the
emergence of irrational behaviors in play.</p></li>
<li><p><a href="https://psycnet.apa.org/record/1981-32566-001">From
exploration to play: A cross-sectional study of infant free play
behavior</a> - <strong><em>Developmental Psychology</em></strong>, 1981.
[<a
href="https://scholar.google.com/scholar?cluster=15547331535034599545&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://srcd.onlinelibrary.wiley.com/doi/abs/10.1111/1467-8624.00224">Detecting
Blickets: How Young Children Use Information about Novel Causal Powers
in Categorization and Induction</a> - <strong><em>Children
Development</em></strong>, 2003. [<a
href="https://scholar.google.com/scholar?cluster=9049737233568227380&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://eccl.scripts.mit.edu/papers/bonawitzandschulzseriousfun.pdf">Serious
fun: Preschoolers engage in more exploratory play when evidence is
confounded</a> - <strong><em>Developmental Psychology</em></strong>,
2007. [<a
href="https://scholar.google.com/scholar?cluster=3033619407322882147&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://stahla.pages.tcnj.edu/files/2015/08/Stahl_Feigenson_Science_2015.pdf">Observing
the unexpected enhances infants’ learning and exploration</a> -
<strong><em>Science</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?start=10&amp;hl=en&amp;as_sdt=0,5&amp;cluster=9247917261616759689">All
Versions</a>].</p></li>
<li><p><a href="https://psycnet.apa.org/record/2008-12114-008">Word,
thought, and deed: the role of object categories in children’s inductive
inferences and exploratory play</a> - <strong><em>Developmental
Psychology</em></strong>, 2009. [<a
href="https://scholar.google.com/scholar?cluster=13947689064550390312&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0010027711000916">Where
science starts: Spontaneous experiments in preschoolers’ exploratory
play</a> - <strong><em>Cognition</em></strong>, 2011. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=16321989770180281706">All
Versions</a>].</p></li>
<li><p><a
href="http://alisongopnik.com/Papers_Alison/Scientific%20Thinking%20in%20young%20Children.pdf">Scientific
thinking in young children: Theoretical advances, empirical research,
and policy implications</a> - <strong><em>Science</em></strong>, 2012.
[<a
href="https://scholar.google.com/scholar?cluster=9103846738385460508&amp;hl=en&amp;as_sdt=2005">All
Versions</a>].</p></li>
<li><p><a
href="http://eccl.scripts.mit.edu/papers/Finding%20New%20Facts_%20Thinking%20New%20Thoughts.pdf">Finding
New Facts; Thinking New Thoughts</a> - <strong><em>Advances in Child
Development and Behavior</em></strong>, 2012. [<a
href="https://scholar.google.com/scholar?hl=en&amp;as_sdt=0%2C5&amp;q=Finding+new+facts%3B+thinking+new+thoughts&amp;btnG=">All
Versions</a>].</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0885201412000445">Theory
learning as stochastic search in the language of thought</a> -
<strong><em>Cognitive Development</em></strong>, 2012. [<a
href="https://scholar.google.com/scholar?cluster=8036476579458645432&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.science.org/doi/abs/10.1126/science.aan2317">Infants
make more attempts to achieve a goal when they see adults persist</a> -
<strong><em>Science</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=2617011825272996810&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://cognitivesciencesociety.org/cogsci20/papers/0716/0716.pdf">Knowing
when to quit: Children consider access to solutions when deciding
whether to persist</a> - <strong><em>CogSci’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=15997297570269958414&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://psyarxiv.com/aq3rp/">Bayesian Models of
Conceptual Development: Learning as Building Models of the World</a> -
<strong><em>Annual Review of Developmental Psychology</em></strong>,
2020. [<a
href="https://scholar.google.com/scholar?cluster=646614032563248495&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://onlinelibrary.wiley.com/doi/full/10.1111/cogs.12765">Sticking
to the Evidence? A Behavioral and Computational Case Study of
Micro-Theory Change in the Domain of Magnetism</a> -
<strong><em>Cognitive Science</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=4409900195679222965&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://junyichu.mit.edu/sites/default/files/documents/2018-05-14%20CogSci%20Final.pdf">Cognitive
pragmatism: Children flexibly choose between facts and conjectures</a> -
<strong><em>CogSci’18</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=6978944437676543728&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://psyarxiv.com/9yra2/">Exploratory play, rational
action, and efficient search</a> - <strong><em>CogSci’20</em></strong>,
2020. [<a
href="https://scholar.google.com/scholar?cluster=17529638197045429028&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://srcd.onlinelibrary.wiley.com/doi/full/10.1111/cdev.13647?saml_referrer">Children
selectively endorse speculative conjectures</a> - <strong><em>Child
Development</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=5672344544260882286&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://psycnet.apa.org/buy/2017-12497-003">Learning
higher-order generalizations through free play: Evidence from 2- and
3-year-old children</a> - <strong><em>Developmental
Psychology</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=4386474921214936914&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://royalsocietypublishing.org/doi/10.1098/rstb.2019.0502">Childhood
as a solution to explore–exploit tensions</a> -
<strong><em>Philosophical Transactions of the Royal Society B:
Biological Sciences</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=11960188575664977017&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.nature.com/articles/s41467-021-23431-2">Children’s
exploratory play tracks the discriminability of hypotheses</a> -
<strong><em>Nature Communications</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=12389351553206792907&amp;hl=en&amp;as_sdt=0,5&amp;as_ylo=2020">All
Versions</a>].</p></li>
<li><p><a
href="https://srcd.onlinelibrary.wiley.com/doi/full/10.1111/j.1467-8624.2010.01499.x?saml_referrer">A
Developmental Perspective on Executive Function</a> - <strong><em>Child
Development</em></strong>, 2010. [<a
href="https://scholar.google.com/scholar?cluster=11347590808138984649&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://journals.sagepub.com/doi/pdf/10.1177/1745691620904771">Rethinking
Executive Function and Its Development</a> - <strong><em>Psychological
Science</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=16570230278367237499&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.harvardlds.org/wp-content/uploads/2017/01/Perception-of-partly-occluded-objects-in-infancy-1.pdf">Perception
of partly occluded objects in infancy</a> - <strong><em>Cognitive
Psychology</em></strong>, 1983. [<a
href="https://scholar.google.com/scholar?cluster=4182861116190610992&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://link.springer.com/article/10.3758/s13428-012-0210-4">Age-of-acquisition
ratings for 30,000 English words</a> - <strong><em>Behavior Research
Methods</em></strong>, 2012. [<a
href="https://scholar.google.com/scholar?cluster=6752414178722956940&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a href="http://crr.ugent.be/archives/806">Project</a>].
A database for age-of-acquisition ratings for over 30k English
words.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="learning-in-the-open-world">Learning in the Open World</h3>
<ul>
<li><p><a
href="https://www.sciencedirect.com/science/article/abs/pii/S002224961730010X">Online
learning of symbolic concepts</a> - <strong><em>Journal of Mathematical
Psychology</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?start=20&amp;hl=en&amp;as_sdt=2005&amp;sciodt=0,5&amp;cites=8036476579458645432&amp;scipsc=">All
Versions</a>].</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&amp;arnumber=8413121">Zero-Shot
Learning—A Comprehensive Evaluation of the Good, the Bad and the
Ugly</a> - <strong><em>IEEE Transactions on Pattern Analysis and Machine
Intelligence</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=11909080239486864961&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A comprehensive review on zero-shot learning.</p></li>
<li><p><a
href="https://www.4paradigm.com/upload/file/20210427/20210427225045_12063.pdf">Generalizing
from a few examples: A survey on few-shot learning</a> - <strong><em>ACM
Computing Survey</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=7932202448069313464&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://ieeexplore.ieee.org/document/7298799">Towards
Open World Recognition</a> - <strong><em>CVPR’15</em></strong>, 2015.
[<a
href="https://scholar.google.com/scholar?cluster=856704237994181529&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The first paper introducing the problem of open-world
recognition.</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&amp;arnumber=7780542">Towards
Open Set Deep Networks</a> - <strong><em>CVPR’16</em></strong>, 2016.
[<a
href="https://scholar.google.com/scholar?cluster=3571743951915089896&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/2007.02519.pdf">In the Wild: From
ML Models to Pragmatic ML Systems</a> -
<strong><em>ICLR’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=15243890330014986346&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A comprehensive review on incremental machine
learning.</p></li>
<li><p><a href="https://arxiv.org/pdf/2002.04108.pdf">Adversarial
Filters of Dataset Biases</a> - <strong><em>ICML’20</em></strong>, 2020.
[<a
href="https://scholar.google.com/scholar?cluster=11617966867048191189&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/2009.01797.pdf">A Wholistic View
of Continual Learning with Deep Neural Networks: Forgotten Lessons and
the Bridge to Active and Open World Learning</a> - 2020. [<a
href="https://scholar.google.com/scholar?cluster=2640432662088551010&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/2011.12216.pdf">Energy-Based
Models for Continual Learning</a> -
<strong><em>NeurIPS’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=7094884707139778576&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://energy-based-model.github.io/Energy-Based-Models-for-Continual-Learning/">Project</a>].</p></li>
<li><p><a
href="https://openaccess.thecvf.com/content_CVPR_2019/papers/Zhou_Learning_to_Learn_Image_Classifiers_With_Visual_Analogy_CVPR_2019_paper.pdf">Learning
to Learn Image Classifiers with Visual Analogy</a> -
<strong><em>CVPR’18</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=6285495755337309034&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/1804.04340v2.pdf">Zero-Shot Object
Detection</a> - <strong><em>ECCV’18</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=2027060030559987993&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/2103.02603v1.pdf">Towards Open
World Object Detection</a> - <strong><em>CVPR’21</em></strong>, 2021.
[<a
href="https://scholar.google.com/scholar?cluster=9715328489246217151&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. [<a
href="https://github.com/JosephKJ/OWOD">Project</a>].</p></li>
<li><p><a
href="https://dl.acm.org/doi/pdf/10.1145/3123266.3123323">Learning to
Recognise Unseen Classes by A Few Similes</a> -
<strong><em>MM’17</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?q=related:FZZr2BK0U6YJ:scholar.google.com/&amp;scioq=Learning+to+Recognise+Unseen+Classes+by+A+Few+Similes&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://proceedings.kr.org/2020/87/kr2020-0087-chen-et-al.pdf">Ontology-guided
Semantic Composition for Zero-Shot Learning</a> -
<strong><em>KR’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=1825132732653262003&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/pdf/2102.07339.pdf">OntoZSL:
Ontology-enhanced Zero-shot Learning</a> -
<strong><em>WWW’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=1042573079110416209&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/2103.00070">Knowledge-aware
Zero-Shot Learning: Survey and Perspective</a> -
<strong><em>IJCAI’21</em></strong> 2021. [<a
href="https://scholar.google.com/scholar?cluster=2596179801089642923&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://ieeexplore.ieee.org/document/8099612">From Red
Wine to Red Tomato: Composition with Context</a> -
<strong><em>CVPR’17</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=6959320578989247472&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://link.springer.com/chapter/10.1007%2F978-3-030-01246-5_11">Attributes
as Operators: Factorizing Unseen Attribute-Object Compositions</a> -
<strong><em>ECCV’18</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=11627198158637727139&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://ieeexplore.ieee.org/document/9010671">Learning
Compositional Representations for Few-Shot Recognition</a> -
<strong><em>CVPR’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=7363445845219257348&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://ieeexplore.ieee.org/document/9156505">Symmetry
and Group in Attribute-Object Compositions</a> -
<strong><em>CVPR’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=16870815556752021056&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2020/file/1010cedf85f6a7e24b087e63235dc12e-Paper.pdf">A
causal view of compositional zero-shot recognition</a> -
<strong><em>NeurIPS’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=2543173389101020482&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://dl.acm.org/doi/10.1145/3394171.3413849">Compositional
Few-Shot Recognition with Primitive Discovery and Enhancing</a> -
<strong><em>MM’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=15817839338790433509&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://ieeexplore.ieee.org/document/9156655">Learning
Unseen Concepts via Hierarchical Decomposition and Composition</a> -
<strong><em>CVPR’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=14161656227038242300&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="learning-with-cognitive-plausibility">Learning with Cognitive
Plausibility</h3>
<ul>
<li><p><a
href="https://en.wikipedia.org/wiki/Accuracy_and_precision">Accuracy and
Precision</a> - <strong><em>Wikipedia</em></strong>. Wikipedia on the
distinctions and the trade-off between accuracy and precision.</p></li>
<li><p><a
href="https://www.annualreviews.org/doi/abs/10.1146/annurev.ps.40.020189.003131">Cognitive
Science: Definition, Status, and Questions</a> - <strong><em>Annual
Review of Psychology</em></strong>, 1989. [<a
href="https://scholar.google.com/scholar?cluster=8549671583307260475&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="http://people.csail.mit.edu/torralba/courses/6.870/papers/Biederman_RBC_1987.pdf">Recognition-by-Components:
A Theory of Human Image Understanding</a> - <strong><em>Psychological
Review</em></strong>, 1987. [<a
href="https://scholar.google.com/scholar?cluster=16522931798979362446&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper on the recognition-by-components
theory.</p></li>
<li><p><a
href="https://www.nature.com/articles/s41586-019-1138-y">Machine
Behaviour</a> - <strong><em>Nature</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=7881171273277686092&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://yzhu.io/publication/dark2020engineering/paper.pdf">Dark,
Beyond Deep: A Paradigm Shift to Cognitive AI with Humanlike Common
Sense</a> - <strong><em>Engineering</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=12292747257300299161&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Yixin Zhu and Song-Chun Zhu’s review on visual
commonsense.</p></li>
<li><p><a
href="https://cims.nyu.edu/~brenden/papers/OrhanEtAl2020NeurIPS.pdf">Self-supervised
Learning Through the eyes of a Child</a> -
<strong><em>NeurIPS’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=5608715260418451299&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Concept learning through near-natural co-occurrence
frequency estimation.</p></li>
<li><p><a href="https://arxiv.org/abs/1910.01442">CLEVRER: CoLlision
Events for Video REpresentation and Reasoning</a> -
<strong><em>ICLR’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=4352064462350202338&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2020/hash/bf15e9bbff22c7719020f9df4badc20a-Abstract.html">BONGARD-LOGO:
A New Benchmark for Human-Level Concept Learning and Reasoning</a> -
<strong><em>NeurIPS’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=9164011458889391917&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://dl.acm.org/doi/10.1145/1143844.1143874">The
relationship between Precision-Recall and ROC curves</a> -
<strong><em>ICML’06</em></strong>, 2006. [<a
href="https://scholar.google.com/scholar?cluster=10708180947310062390&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="http://export.arxiv.org/pdf/2009.08092">Distributional
Generalization: A New Kind of Generalization</a> - 2020. [<a
href="https://scholar.google.com/scholar?cluster=6190621467796247477&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/abs/pii/0010027793900584">Learning
and development in networks: The importance of starting small.</a> -
<strong><em>Cognition</em></strong>, 1993. [<a
href="https://scholar.google.com/scholar?cluster=5133345254007462915&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper on the idea of curriculum
learning.</p></li>
<li><p><a
href="https://www.sciencedirect.com/science/article/pii/S0010027799000311">Language
acquisition in the absence of explicit negative evidence: how important
is starting small?</a> - <strong><em>Cognition</em></strong>, 1999. [<a
href="https://scholar.google.com/scholar?cluster=11813578367725362166&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://dl.acm.org/doi/pdf/10.1145/1553374.1553380">Curriculum
Learning</a> - <strong><em>ICML’09</em></strong>, 2009. [<a
href="https://scholar.google.com/scholar?cluster=8740915934335425405&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper applying the idea of curriculum
learning to machine learning.</p></li>
<li><p><a href="https://ieeexplore.ieee.org/document/6126279">Parsing
video events with goal inference and intent prediction</a> -
<strong><em>ICCV’11</em></strong>, 2011. [<a
href="https://scholar.google.com/scholar?cluster=5979196784405021658&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://ieeexplore.ieee.org/document/6751387">Inferring
“Dark Matter” and “Dark Energy” from Videos</a> -
<strong><em>ICCV’13</em></strong>, 2013. [<a
href="https://scholar.google.com/scholar?cluster=3467068307444498624&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. The original paper on latent state discovery from
videos.</p></li>
<li><p><a
href="https://openaccess.thecvf.com/content_CVPR_2019/papers/Shi_Explainable_and_Explicit_Visual_Reasoning_Over_Scene_Graphs_CVPR_2019_paper.pdf">Explainable
and Explicit Visual Reasoning over Scene Graphs</a> -
<strong><em>CVPR’19</em></strong>, 2019. [<a
href="https://scholar.google.com/scholar?cluster=8517395712319798436&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2021/hash/4c26774d852f62440fc746ea4cdd57f6-Abstract.html">Attention
over Learned Object Embeddings Enables Complex Visual Reasoning</a> -
<strong><em>NeurIPS’21</em></strong>, 2021. [<a
href="https://scholar.google.com/scholar?cluster=127829313460149801&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://papers.NeurIPS.cc/paper/2013/file/9aa42b31882ec039965f3c4923ce901b-Paper.pdf">Distributed
Representations of Words and Phrases and their Compositionality</a> -
<strong><em>NeurIPS’13</em></strong>, 2013. [<a
href="https://scholar.google.com/scholar?cluster=2410615501856807729&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&amp;arnumber=9197172">Motion
Reasoning for Goal-Based Imitation Learning</a> -
<strong><em>ICRA’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=7519230802512388210&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://openaccess.thecvf.com/content_CVPR_2020/papers/Ji_Action_Genome_Actions_As_Compositions_of_Spatio-Temporal_Scene_Graphs_CVPR_2020_paper.pdf">Action
Genome: Actions as Compositions of Spatio-temporal Scene Graphs</a> -
<strong><em>CVPR’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=388714326304810525&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://proceedings.neurips.cc/paper/2020/file/64dcf3c521a00dbb4d2a10a27a95a9d8-Paper.pdf">Refactoring
Policy for Compositional Generalizability using Self-Supervised Object
Proposals</a> - <strong><em>NeurIPS’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=2255457416066730255&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://openaccess.thecvf.com/content_CVPR_2020/papers/Materzynska_Something-Else_Compositional_Action_Recognition_With_Spatial-Temporal_Interaction_Networks_CVPR_2020_paper.pdf">Something-Else:
Compositional Action Recognition with Spatial-Temporal Interaction
Networks</a> - <strong><em>CVPR’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=17469863154797360929&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://openaccess.thecvf.com/content_CVPR_2020/papers/Zhang_Putting_Visual_Object_Recognition_in_Context_CVPR_2020_paper.pdf">Putting
visual object recognition in context</a> -
<strong><em>CVPR’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=6207193649298787857&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/2106.13884">Multimodal Few-Shot
Learning with Frozen Language Models</a> - 2021. [<a
href="https://scholar.google.com/scholar?cluster=16154696122208258147&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&amp;arnumber=5206772">Describing
Objects by their Attributes</a> - <strong><em>CVPR’09</em></strong>,
2009. [<a
href="https://scholar.google.com/scholar?cluster=6853730684095116174&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://arxiv.org/abs/2108.07783">Panoramic Learning
with A Standardized Machine Learning Formalism</a> - 2021. [<a
href="https://scholar.google.com/scholar?cluster=14222434793711614257&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://psycnet.apa.org/record/1996-10319-001">Graininess of
judgment under uncertainty: An accuracy-informativeness trade-off</a> -
<strong><em>Journal of Experimental Psychology</em></strong>, 1995. [<a
href="https://scholar.google.com/scholar?cluster=15366302654259490472&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://openreview.net/forum?id=GFsU8a0sGB">Federated
Learning via Posterior Averaging: A New Perspective and Practical
Algorithms</a> - <strong><em>ICLR’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=2486025806014234529&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.biorxiv.org/content/10.1101/2022.01.29.478330v2.abstract">Interplay
between rule learning and rule switching in a perceptual categorization
task</a> - 2022. [<a
href="https://scholar.google.com/scholar?cluster=7461559646167397406&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<!--
### Tasks & Environments
#### Dataset Aggregation
* [A Dataset and Architecture for Visual Reasoning with a Working Memory](https://link.springer.com/chapter/10.1007%2F978-3-030-01249-6_44) - ***ECCV'18***, 2018. [[Project](https://github.com/google/cog)].
* [PHYRE: A New Benchmark for Physical Reasoning](https://research.fb.com/wp-content/uploads/2019/08/PHYRE-A-New-Benchmark-for-Physical-Reasoning-v4.pdf) - ***NeurIPS'19***, 2019.
* [CATER: A diagnostic dataset for Compositional Actions & TEmporal Reasoning](https://openreview.net/forum?id=HJgzt2VKPB) - ***ICLR'20***, 2020. [[Project](https://rohitgirdhar.github.io/CATER/)].
* [CausalWorld: A Robotic Manipulation Benchmark for Causal Structure and Transfer Learning](https://arxiv.org/abs/2010.04296), 2020.
#### Embodied AI Environment
* [ThreeDWorld](http://www.threedworld.org/) - ***MIT-IBM***. [[Paper](https://arxiv.org/abs/2007.04954)].
* [Rearrangement: A Challenge for Embodied AI](https://arxiv.org/pdf/2011.01975.pdf), 2020.
* [iGibson](http://svl.stanford.edu/igibson/) - ***Stanford***. [[Paper](https://ieeexplore.ieee.org/document/8954627)].
* [AI2-THOR](https://ai2thor.allenai.org/ithor) - ***Allen Institute***. [[Paper](https://arxiv.org/abs/1712.05474)].
* [Robo-THOR](https://ai2thor.allenai.org/robothor) - ***Allen Institute***. [[Paper](https://arxiv.org/abs/2004.06799)].
* [Manipula-THOR](https://ai2thor.allenai.org/manipulathor) - ***Allen Institute***. [[Paper](https://arxiv.org/abs/2104.11213)].
* [RLBench](https://sites.google.com/view/rlbench) - ***Imperial College***. [[Paper](https://ieeexplore.ieee.org/document/9001253)].
#### First-Person Vision
* [First-Person Vision](https://ieeexplore.ieee.org/document/6232429) - ***Proceedings of the IEEE***, 2012.
* [The Evolution of First Person Vision Methods: A Survey](https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7055926) - ***Trans. CSVT***, 2015.
* [Understanding the Nature of First-Person Videos: Characterization and Classification using Low-Level Features](http://vijaychan.github.io/Publications/2014%20CVPR%20Workshop%20-%20Understanding%20the%20Nature%20of%20First-Person%20Videos.pdf) - ***CVPR'14***, 2014.
* [Pooled Motion Features for First-Person Videos](https://openaccess.thecvf.com/content_cvpr_2015/papers/Ryoo_Pooled_Motion_Features_2015_CVPR_paper.pdf) - ***CVPR'15***, 2015.
* [Actor and Observer: Joint Modeling of First and Third-Person Videos](https://openaccess.thecvf.com/content_cvpr_2018/papers/Sigurdsson_Actor_and_Observer_CVPR_2018_paper.pdf) - ***CVPR'18***, 2018.
* [Forecasting Human-Object Interaction: Joint Prediction of Motor Attention and Actions in First Person Video](https://link.springer.com/chapter/10.1007/978-3-030-58452-8_41) - ***ECCV'20***, 2020.
* [Rolling-Unrolling LSTMs for Action Anticipation from First-Person Video](https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9088213) - ***Trans. PAMI***, 2020.
* [View-Action Representation Learning for Active First-Person Vision](https://ieeexplore.ieee.org/document/9064828) - ***Trans. CSVT***, 2021.
* [Design and Use Paradigms for Gazebo, An Open-Source Multi-Robot Simulator](https://ieeexplore.ieee.org/abstract/document/1389727) - ***IROS'04***, 2004. [[Project](http://gazebosim.org/)].
* [ViZDoom: A Doom-based AI Research Platform for Visual Reinforcement Learning](https://arxiv.org/pdf/1605.02097v2.pdf) - ***CIG'16***, 2016. [[Project](http://vizdoom.cs.put.edu.pl/)].
* [Is First Person Vision Challenging for Object Tracking? The TREK-100 Benchmark Dataset](https://arxiv.org/abs/2011.12263), 2020.
* **Visual Experience Database** [[Project](http://visualexperiencedatabase.org/research.html)]. [[Publications](http://visualexperiencedatabase.org/publications.html)].
#### Abstract Reasoning Challenge
* [On the Measure of Intelligence](https://arxiv.org/pdf/1911.01547.pdf) - ***Google Research***, 2019.
* [Abstract Reasoning Challenge](https://www.kaggle.com/c/abstraction-and-reasoning-challenge/)
#### AI Birds Challenge
* [AI-Birds](https://aibirds.org) - ***IJCAI***.
* [Hi-Phy: A Benchmark for Hierarchical Physical Reasoning](https://openreview.net/forum?id=AcL1ORzw0Nf), 2021.
#### Minecraft
* [Mining Learning and Crafting Scientific Experiments: A Literature Review on the Use of Minecraft in Education and Research](https://eric.ed.gov/?id=EJ1097278) - ***Journal on Eduction Technology & Society***, 2016.
##### Malmo Platform for Minecraft AI
* [The Malmo Platform for Artificial Intelligence Experimentation](https://www.microsoft.com/en-us/research/publication/malmo-platform-artificial-intelligence-experimentation/) ***IJCAI'16***, 2016.
* [[Malmo](https://github.com/Microsoft/malmo#getting-started)].
* [[Malmo-env](https://github.com/Microsoft/malmo/tree/master/MalmoEnv)].
* [[Malmo-Tutorials](https://microsoft.github.io/malmo/0.17.0/Python_Examples/Tutorial.pdf)].
* [[MineRL](https://minerl.io/)].
* [[MarLo Challenge 2018](https://github.com/crowdAI/marLo)].
##### **Artificial Intelligence**
* [Multi-task curriculum learning in a complex, visual, hard-exploration domain: Minecraft](https://arxiv.org/abs/2106.14876), 2021.
* [Learning to execute instructions in a Minecraft dialogue](https://www.aclweb.org/anthology/2020.acl-main.232/) - ***ACL'20***, 2020.
* [Collaborative Dialogue in Minecraft](https://www.aclweb.org/anthology/P19-1537.pdf) - ***ACL'19***, 2019.
* [Learning Skill Hierarchies from Predicate Descriptions and Self-Supervision](http://web.mit.edu/tslvr/www/papers/genplan20_camera_ready.pdf) - ***AAAI GenPlan Workshop***, 2020.
* [AMRL: Aggregated Memory for Reinforcement Learning](https://openreview.net/pdf?id=Bkl7bREtDr) - ***ICLR'20***, 2020.
* [MineRL: A Large-Scale Dataset of Minecraft Demonstrations](https://www.ijcai.org/Proceedings/2019/0339.pdf) ***IJCAI'19***, 2019. [[2020 Competition](https://arxiv.org/abs/2106.03748)].
* [Design Mining for Minecraft Architecture](http://www.cs.cornell.edu/~eland/papers/aiide2018.pdf) - ***AAAI'18***, 2018.
* [Adaptive Agents in Minecraft: A Hybrid Paradigm for Combining Domain Knowledge with Reinforcement Learning](https://link.springer.com/chapter/10.1007%2F978-3-319-71679-4_6) - ***AAMAS'17***, 2017.
* [Asynchronous Data Aggregation for Training End to End Visual Control Networks](https://www.microsoft.com/en-us/research/wp-content/uploads/2017/05/fp185-monfort-1.pdf) - ***AAMAS'17***, 2017.
* [A Deep Hierarchical Approach to Lifelong Learning in Minecraft](https://aaai.org/ocs/index.php/AAAI/AAAI17/paper/view/14630/13950) - ***AAAI'17***, 2017.
* [Modular Multitask Reinforcement Learning with Policy Sketches](http://proceedings.mlr.press/v70/andreas17a.html) - ***ICML'17***, 2017.
* [Control of memory, active perception, and action in minecraft](http://proceedings.mlr.press/v48/oh16.pdf) - ***ICML'16***, 2016.
* [Learning Behavior from Demonstration in Minecraft via Symbolic Similarity Measures](fdg2015.org/papers/fdg2015_paper_11.pdf) - ***FDG'15***, 2015.
##### **Cognitive Science**
* [How Players Speak to an Intelligent GameCharacter Using Natural Language Messages](http://todigra.org/index.php/todigra/article/view/88/139) - ***DiGRA***, 2018.
* [Minecraft as a Generative Platform for Analyzing and Practicing Spatial Reasoning](https://link.springer.com/chapter/10.1007%2F978-3-030-57983-8_22) - ***Spatial Cognition'20***, 2020.
* [Generative Design in Minecraft: Chronicle Challenge](http://computationalcreativity.net/iccc2019/papers/iccc19-lbp-7.pdf) - ***ICCC'20***, 2020.
* [Minecraft as a Platform for Project-Based Learning in AI](https://aaai.org/ojs/index.php/AAAI/article/view/7070) - ***AAAI'20***, 2020.
* [MC-Saar-Instruct: a Platform for Minecraft Instruction Giving Agents](https://www.aclweb.org/anthology/2020.sigdial-1.7.pdf) - ***SIGDial'20***, 2020.
*[Back to Top](#c)-->
<h2 id="academic-tools">Academic Tools</h2>
<h3 id="courses">Courses</h3>
<ul>
<li><p><a
href="https://cbmm.mit.edu/education/courses/computational-cognitive-science">Computational
Cognitive Science Courses</a> - <strong><em>MIT</em></strong>. Courses
on computational cognitive science from MIT, Harvard, and
Stanford.</p></li>
<li><p><a
href="https://people.csail.mit.edu/asolar/SynthesisCourse/index.htm">Introduction
to Program Synthesis</a> - <strong><em>MIT</em></strong>. Armando
Solar-Lezama’s elementary course on program synthesis.</p></li>
<li><p><a href="https://web.mit.edu/6.001/6.037/">Structure and
Interpretation of Computer Programs</a> - <strong><em>MIT</em></strong>.
[<a href="https://web.mit.edu/6.001/6.037/sicp.pdf">Book: SICP</a>]. [<a
href="https://scholar.google.com/scholar?cluster=7488066943428166450&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Classic course on applying structural, procedural, and
meta-linguistic abstraction to solve computational problems.</p></li>
<li><p><a
href="https://faculty.ksu.edu.sa/sites/default/files/rosen_discrete_mathematics_and_its_applications_7th_edition.pdf">Discrete
Mathematics and Its Applications</a>. Classic course on basic discrete
mathematics, including matheatical logic, set theory, graph theory,
formal language (and automata), basic number theory (e.g., counting),
and other related topics.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="programming">Programming</h3>
<ul>
<li><a href="https://probmods.org/">Probabilistic Models of
Cognition</a> - <strong><em>MIT</em></strong>. The probabilistic
approach to cognitive science, which models learning and reasoning as
inference in complex probabilistic models.</li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="paper-writing">Paper Writing</h3>
<ul>
<li><p><a href="LaTex/config.sty">LaTex Configuration</a> -
<strong><em>LaTex</em></strong>. LaTex template for configuration file
with elegant reference style (gray-colored reference, page backward
reference).</p></li>
<li><p><a href="BibTex/references_header.bib">BibTex Template</a> -
<strong><em>BibTex</em></strong>. BibTex template for including
abbreviations of journals and conferences in AI, Mathematics, and
Cognitive Sciences.</p></li>
<li><p><a href="https://www.biorender.com/">bioRender</a> -
<strong><em>bioRender</em></strong>. Create professional science figures
in minutes by browsing thousands of pre-made icons and templates from
more than 30 fields of life sciences.</p></li>
<li><p><a
href="https://www.nature.com/documents/nature-summary-paragraph.pdf">How
to construct a Nature summary paragraph</a> -
<strong><em>Nature</em></strong>. Nature official guidelines for
composing abstracts.</p></li>
<li><p><a href="https://www.nature.com/articles/d41586-020-03422-x">How
to write a superb literature review</a> -
<strong><em>Nature</em></strong>, 2020. Nature speaks to old hands and
first timers about the work they did to make their reviews
sing.</p></li>
<li><p><a
href="https://www.nature.com/scitable/topicpage/scientific-papers-13815490/">Scientific
Papers</a> - <strong><em>Nature</em></strong>. Nature guidance on
writing scientific papers.</p></li>
<li><p><a
href="https://www.cs.mcgill.ca/~jpineau/ReproducibilityChecklist.pdf">The
Machine Learning Reproducibility Checklist</a> - <strong><em>McGill
University</em></strong>. Guidelines for introducing a machine learning
algorithm with guarantee of reproducibility.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="paper-reading">Paper Reading</h3>
<ul>
<li><p><a
href="https://www.cs.uni-potsdam.de/bs/teaching/docs/courses/ss2020/scn/material/p83-keshavA.pdf">How
to Read a Paper</a> - <strong><em>ACM SIGCOMM Computer Communication
Review</em></strong>, 2007. [<a
href="https://scholar.google.com/scholar?cluster=7234542241721187587&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A comprehensive tutorial on reading scientific
papers.</p></li>
<li><p><a
href="https://www.science.org/content/article/how-seriously-read-scientific-paper">How
to (seriously) read a scientific paper</a> -
<strong><em>Science</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?hl=en&amp;as_sdt=0%2C5&amp;q=How+to+%28seriously%29+read+a+scientific+paper&amp;btnG=">All
Versions</a>]. Science interview on reading scientific papers.</p></li>
<li><p><a href="https://www.nature.com/articles/nature.2017.21751">It’s
not just you: science papers are getting harder to read</a> -
<strong><em>Nature</em></strong>, 2017. [<a
href="https://scholar.google.com/scholar?cluster=4409814498614719804&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Nature perspective on reading scientific papers.</p></li>
<li><p><a
href="https://be.mit.edu/sites/default/files/documents/HowToReadAScientificPaper.pdf">How
to navigate a scientific paper with time constraints: a graphics
approach</a> - <strong><em>MIT</em></strong>. MIT guidance on strategies
for reading papers given different time constraints.</p></li>
<li><p><a href="https://textvis.lnu.se/">Text Visualization Browser</a>
- <strong><em>ISOVIS group</em></strong>, 2015. [<a
href="https://cs.lnu.se/isovis/pubs/docs/kucher-pacificvis15-postprint.pdf">Paper</a>].
[<a
href="https://scholar.google.com/scholar?cluster=7000995325728444282&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A Hub of Text Visualization Techniques.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="literature-management">Literature Management</h3>
<ul>
<li><p><a
href="https://www.science.org/content/article/how-keep-scientific-literature">How
to keep up with the scientific literature</a> -
<strong><em>Science</em></strong>, 2016. Science interview on organizing
scientific papers.</p></li>
<li><p><a href="https://www.nature.com/articles/nj7612-457a">Scientific
literature: Information overload</a> - <strong><em>Nature</em></strong>,
2016. [<a
href="https://scholar.google.com/scholar?cluster=9898832432826237365&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Perspective on handling overloaded information from
scientific literature.</p></li>
<li><p><a
href="https://www.microsoft.com/en-us/research/project/microsoft-academic-graph/">Microsoft
Academic Graph</a> - <strong><em>Microsoft Research</em></strong>.
Heterogeneous graph containing scientific publication records, citation
relationships between those publications, as well as authors,
institutions, journals, conferences, and fields of study.</p></li>
<li><p><a
href="http://sonyis.me/paperpdf/Microsoft%20Academic%20Graph%20WWW%202015.pdf">An
Overview of Microsoft Academic Service (MAS) and Applications</a> -
<strong><em>WWW’15</em></strong>, 2015. [<a
href="https://scholar.google.com/scholar?cluster=9075899176667058496&amp;hl=en&amp;as_sdt=0,5">All
Versios</a>]. Original paper on Microsoft Academic Graph.</p></li>
<li><p><a
href="https://blogs.lse.ac.uk/impactofsocialsciences/2021/05/27/goodbye-microsoft-academic-hello-open-research-infrastructure/">Goodbye,
Microsoft Academic – Hello, open research infrastructure?</a> -
<strong><em>LSE Impact Blog</em></strong>, 2021. An interpretation of
Microsoft’s strategy on research infrastructure.</p></li>
<li><p><a href="https://www.semanticscholar.org/">Semantic Scholar</a> -
<strong><em>Allen Institute for AI Research</em></strong>. AI-powered
scientific literature research tool.</p></li>
<li><p><a href="https://aclanthology.org/N18-3011/">Construction of the
Literature Graph in Semantic Scholar</a> -
<strong><em>NAACL’18</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=5500969515339734950&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. Semantic Scholar with extracting feature and metadata
from raw paper data.</p></li>
<li><p><a href="https://aclanthology.org/2020.acl-main.447/">S2ORC: The
Semantic Scholar Open Research Corpus</a> -
<strong><em>ACL’20</em></strong>, 2020. [<a
href="https://scholar.google.com/scholar?cluster=11978464475399626925&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. An open corpus of academic papers released by Semantic
Scholar.</p></li>
<li><p><a href="https://www.litmaps.com/">Litmaps</a> -
<strong><em>Litmap Ltd</em></strong>. For interactive literature map
construction and linked document management.</p></li>
<li><p><a href="https://www.vosviewer.com/">VOSviewer</a> -
<strong><em>Leiden University</em></strong>. For constructing and
visualizing bibliometric networks.</p></li>
<li><p><a href="https://www.stateoftheart.ai/">StateOfTheArt.AI</a> -
<strong><em>StateOfTheArtAI</em></strong>. For tracking, collecting and
visualizing the development of AI research.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="knowledge-management">Knowledge Management</h3>
<ul>
<li><p><a
href="https://www.loc.gov/aba/publications/FreeLCC/freelcc.html">Library
of Congress Classification</a> - <strong><em>Library of
Congress</em></strong>. Classification system of USA (PDF
only).</p></li>
<li><p><a href="http://cct.nlc.cn/">Chinese Library Classification</a> -
<strong><em>National Library of China</em></strong>. Classification
system of P. R. China (online user interface in Chinese). [<a
href="https://www.isko.org/cyclo/clc">English introduction at ISKO</a>].
[<a
href="https://en.wikipedia.org/wiki/Chinese_Library_Classification">Wikipedia-EN</a>].</p></li>
<li><p><a
href="https://rvk.uni-regensburg.de/regensburger-verbundklassifikation-online">DDC
at German National Library</a> - <strong><em>Deutsche National
Bibliothek</em></strong>. Deway Decimal Classification (DDC) based
classification system of Germany (online user interface). [<a
href="https://www.dnb.de/EN/Professionell/DDC-Deutsch/DDCinDNB/ddcindnb_node.html">DNB
Website</a>].</p></li>
<li><p><a
href="https://www.ndl.go.jp/jp/data/catstandards/classification_subject/ndlc.html">National
Dite Library Classification</a> - <strong><em>National Diet Library of
Japan</em></strong>. Classification system of Japan (PDF only).</p></li>
<li><p><a
href="https://en.wikipedia.org/wiki/List_of_Dewey_Decimal_classes">DDC
at OCLC (Wikipedia)</a> - <strong><em>Online Computer Library Center
(OCLC)</em></strong>. [<a href="https://www.oclc.org/en/home.html">OCLC
Website</a>]. [<a
href="https://www.oclc.org/content/dam/oclc/dewey/versions/print/intro.pdf">Introduction
to DDC</a>]. [<a
href="https://www.oclc.org/content/dam/oclc/webdewey/help/full_manual.pdf">DDC
Manual</a>]. Dewey Decimal Classification (DDC) system for worldwide
library resouce construction. [<a
href="https://www.oclc.org/content/dam/oclc/webdewey/help/000.pdf">DDC
Class 000 (PDF only)</a>]. [<a
href="https://www.oclc.org/content/dam/oclc/webdewey/help/100.pdf">DDC
Class 100 (PDF only)</a>]. [<a
href="https://www.oclc.org/content/dam/oclc/webdewey/help/200.pdf">DDC
Class 200 (PDF only)</a>]. [<a
href="https://www.oclc.org/content/dam/oclc/webdewey/help/300.pdf">DDC
Class 300 (PDF only)</a>]. [<a
href="https://www.oclc.org/content/dam/oclc/webdewey/help/400.pdf">DDC
Class 400 (PDF only)</a>]. [<a
href="https://www.oclc.org/content/dam/oclc/webdewey/help/500.pdf">DDC
Class 500 (PDF only)</a>]. [<a
href="https://www.oclc.org/content/dam/oclc/webdewey/help/600.pdf">DDC
Class 600 (PDF only)</a>]. [<a
href="https://www.oclc.org/content/dam/oclc/webdewey/help/700.pdf">DDC
Class 700 (PDF only)</a>]. [<a
href="https://www.oclc.org/content/dam/oclc/webdewey/help/800.pdf">DDC
Class 800 (PDF only)</a>]. [<a
href="https://www.oclc.org/content/dam/oclc/webdewey/help/900.pdf">DDC
Class 900 (PDF only)</a>].</p></li>
<li><p><a
href="https://en.wikipedia.org/wiki/Knowledge_organization">Knowledge
organization</a> - <strong><em>Wikipedia</em></strong>. Wikipedia on
knowledge organization methods.</p></li>
<li><p><a href="https://zettelkasten.de/">The Zettelkasten Method</a> -
<strong><em>Bielefeld University</em></strong>. Relating ideas in graphs
and multi-labels.</p></li>
<li><p><a
href="https://en.wikipedia.org/wiki/Zettelkasten">Zettelkasten</a> -
<strong><em>Wikipedia</em></strong>. Wikipedia on the Zettelkasten
method.</p></li>
<li><p><a href="https://roamresearch.com/">Roam Research</a> -
<strong><em>Roam Research</em></strong>. For linked document management,
visualization, and sharing.</p></li>
<li><p><a href="https://foambubble.github.io/foam/">Foam</a> -
<strong><em>Foambubble</em></strong>. For linked document management,
visualization, and sharing, opensourced softward built on
VSCode.</p></li>
<li><p><a href="https://www.buildingasecondbrain.com/">Building a Second
Brain</a> - <strong><em>Forte Labs, LLC</em></strong>. Connecting ideas
in graphs.</p></li>
<li><p><a href="https://www.zotero.org/">Zotero</a> -
<strong><em>Digital Scholar</em></strong>. For reference management to
manage bibliographic data and research related materials.</p></li>
<li><p><a
href="https://pdfs.semanticscholar.org/88f8/fa9dfbc0c2b296758dd932b871917c5c775a.pdf%C2%A0">Niklas
Luhmann’s Card Index: Thinking Tool, Communication Partner, Publication
Machine</a> - <strong><em>Forgetting Machines: Knowledge Management
Evolution in Early Modern Europe, Brill</em></strong>, 2016. [<a
href="https://scholar.google.com/scholar?cluster=1786807670077004336&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://culturemachine.net/wp-content/uploads/2019/01/373-604-1-PB.pdf">The
card index as creativity machine</a> - <strong><em>Culture
Machine</em></strong>, 2010. [<a
href="https://scholar.google.com/scholar?cluster=9767873312286889264&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://www.researchgate.net/profile/Alberto-Cevolini/publication/328624186_Where_Does_Niklas_Luhmann%27s_Card_Index_Come_From/links/609f818e299bf147699a401d/Where-Does-Niklas-Luhmanns-Card-Index-Come-From.pdf">Where
Does Niklas Luhmann’s Card Index Come From?</a> - <strong><em>Erudition
and the Republic of Letters</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=8279465066043884141&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>]. A simplified introduction on Luhmann’s
Zettelkasten.</p></li>
<li><p><a
href="https://www.uni-bielefeld.de/fakultaeten/soziologie/forschung/luhmann-archiv/pdf/jschmidt_niklas-luhmanns-card-index_-sociologica_2018_12-1.pdf">Niklas
Luhmann’s Card Index: The Fabrication of Serendipity</a> -
<strong><em>Sociologica</em></strong>, 2018. [<a
href="https://scholar.google.com/scholar?cluster=12440286698665929622&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://luhmann.surge.sh/communicating-with-slip-boxes">Communicating
with Slip Boxes</a> - 2019. [<a
href="https://scholar.google.com/scholar?hl=en&amp;as_sdt=0%2C5&amp;q=Communicating+with+slip+boxes+luhmann&amp;btnG=">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h2 id="institute-researcher">Institute &amp; Researcher</h2>
<h3 id="mit">MIT</h3>
<ul>
<li><p><a href="https://cbmm.mit.edu/">Center for Brains, Minds and
Machines (CBMM)</a> - <strong><em>MIT</em></strong>.</p></li>
<li><p><a href="https://cocosci.mit.edu/josh">Josh Tenenbaum</a> -
<strong><em>Department of Brain and Cognitive Sciences, CSAIL,
MIT</em></strong>, <a href="https://cocosci.mit.edu/">Computational
Cognitive Science Group (CoCoSci Group)</a> -
<strong><em>MIT</em></strong>.</p></li>
<li><p><a href="https://saxelab.mit.edu/people/rebecca-saxe">Rebecca
Saxe</a> - <strong><em>Department of Brain and Cognitive Sciences,
MIT</em></strong>, <a href="https://saxelab.mit.edu/">Social Cognitive
Neuroscience Laboratory (SaxeLab)</a> -
<strong><em>MIT</em></strong>.</p></li>
<li><p><a href="https://cbmm.mit.edu/about/people/schulz">Laura
Schulz</a> - <strong><em>Department of Brain and Cognitive Sciences,
MIT</em></strong>, <a href="https://eccl.mit.edu/">Early Childhood
Cognition Lab</a> - <strong><em>MIT</em></strong>.</p></li>
<li><p><a href="https://people.csail.mit.edu/lpk/">Leslie Kaelbling</a>
- <strong><em>Department of Electrical Engineering and Computer Science,
CSAIL, MIT</em></strong>, <a href="https://lis.csail.mit.edu/">The
Learning &amp; Intelligent Systems Group</a> -
<strong><em>MIT</em></strong>.</p></li>
<li><p><a href="https://people.csail.mit.edu/asolar/">Armando
Solar-Lezama</a> - <strong><em>Department of Electrical Engineering and
Computer Science, CSAIL, MIT</em></strong>, <a
href="http://groups.csail.mit.edu/cap/">Computer-Aided Programming
Group</a> - <strong><em>MIT</em></strong>.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="stanford">Stanford</h3>
<ul>
<li><p><a href="https://profiles.stanford.edu/fei-fei-li">Li Fei-Fei</a>
- <strong><em>Computer Science Department, Human-Centered AI Institute,
Stanford</em></strong>, <a href="https://svl.stanford.edu/">Stanford
Vision and Learning Lab</a> -
<strong><em>Stanford</em></strong>.</p></li>
<li><p><a href="https://cocolab.stanford.edu/ndg.html">Noah Goodman</a>
- <strong><em>Department of Psychology, Computer Science Department,
Stanford</em></strong>, <a
href="https://cocolab.stanford.edu/">Computation &amp; Cognition Lab
(CoCoLab)</a> - <strong><em>Stanford</em></strong>.</p></li>
<li><p><a href="https://web.stanford.edu/~mcfrank/">Michael Frank</a> -
<strong><em>Department of Psychology, Stanford</em></strong>, <a
href="http://langcog.stanford.edu/">The Stanford Language and Cognition
Lab</a> - <strong><em>Stanford</em></strong>.</p></li>
<li><p><a
href="https://cicl.stanford.edu/member/tobias_gerstenberg/">Tobias
Gerstenberg</a> - <strong><em>Department of Psychology,
Stanford</em></strong>, <a href="https://cicl.stanford.edu/">Causality
in Cognition Lab (CICL)</a> -
<strong><em>Stanford</em></strong>.</p></li>
<li><p><a href="http://ai.stanford.edu/~cbfinn/">Chelsea Finn</a> -
<strong><em>Computer Science Department, Stanford</em></strong>, <a
href="https://irislab.stanford.edu/">Intelligence through Robotic
Interaction at Scale (IRIS Group)</a> -
<strong><em>Stanford</em></strong>.</p></li>
<li><p><a href="https://comm.stanford.edu/faculty-bailenson/">Jeremy
Bailenson</a> - <strong><em>Department of Communication,
Stanford</em></strong>, <a href="https://stanfordvr.com/">Virtual Human
Interaction Lab (VHIL)</a> -
<strong><em>Stanford</em></strong>.</p></li>
<li><p><a href="https://jiajunwu.com/">Jiajun Wu</a> -
<strong><em>Computer Science Department,
Stanford</em></strong>.</p></li>
<li><p><a href="https://profiles.stanford.edu/judith-fan">Judith Fan</a>
- <strong><em>Department of Psychology, Stanford</em></strong>, <a
href="https://cogtoolslab.github.io/">Cognitive Tools Lab</a> -
<strong><em>Stanford</em></strong>.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="princeton">Princeton</h3>
<ul>
<li><p><a href="https://psych.princeton.edu/person/tania-lombrozo">Tania
Lombrozo</a> - <strong><em>Department of Psychology,
Princeton</em></strong>, <a
href="https://cognition.princeton.edu/">Concepts &amp; Cognition Lab</a>
- <strong><em>Princeton</em></strong>.</p></li>
<li><p><a href="https://cocosci.princeton.edu/tom/index.php">Thomas
Griffiths</a> - <strong><em>Department of Psychology, Department of
Computer Science, Princeton</em></strong>, <a
href="https://cocosci.princeton.edu/index.php">Computational Cognitive
Science Lab</a> - <strong><em>Princeton</em></strong>.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="harvard">Harvard</h3>
<ul>
<li><p><a
href="https://psychology.fas.harvard.edu/people/elizabeth-s-spelke">Elizabeth
Spelke</a> - <strong><em>Department of Psychology,
Harvard</em></strong>, <a href="https://www.harvardlds.org/">Harvard
Laboratory for Developmental Studies</a> -
<strong><em>Harvard</em></strong>.</p></li>
<li><p><a href="https://www.tomerullman.org/">Tomer Ullman</a> -
<strong><em>Department of Psychology, Harvard</em></strong>, <a
href="https://cocodev.fas.harvard.edu/">Computation, Cognition, and
Development Lab (CoCoDev)</a> -
<strong><em>Harvard</em></strong>.</p></li>
<li><p><a
href="https://psychology.fas.harvard.edu/people/samuel-j-gershman">Samuel
Gershman</a> - <strong><em>Department of Psychology,
Harvard</em></strong>, <a href="https://gershmanlab.com/">Computational
Cognitive Neuroscience Lab (CCN Lab)</a> -
<strong><em>Harvard</em></strong>.</p></li>
<li><p><a
href="https://psychology.fas.harvard.edu/people/fiery-cushman">Fiery
Cushman</a> - <strong><em>Department of Psychology,
Harvard</em></strong>, <a
href="https://cushmanlab.fas.harvard.edu/">Moral Psychology Research
Lab</a> - <strong><em>Harvard</em></strong>.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="ucla">UCLA</h3>
<ul>
<li><p><a href="http://vcla.stat.ucla.edu/">Center for Vision,
Cognition, Learning and Autonomy (VCLA)</a> - <strong><em>Department of
Statistics, UCLA</em></strong>.</p></li>
<li><p><a href="http://www.stat.ucla.edu/~ywu/">Ying Nian Wu</a> -
<strong><em>Department of Statistics, UCLA</em></strong>.</p></li>
<li><p><a href="http://www.stat.ucla.edu/~taogao/Taogao.html">Tao
Gao</a> - <strong><em>Department of Statistics, Department of
Psychology, UCLA</em></strong>, <a
href="http://www.stat.ucla.edu/~taogao/index.html">Visual Intelligence
Lab</a> - <strong><em>UCLA</em></strong>.</p></li>
<li><p><a
href="https://www.psych.ucla.edu/faculty/page/hongjing">Hongjing Lu</a>
- <strong><em>Department of Psychology, Department of Statistics,
UCLA</em></strong>, <a href="http://cvl.psych.ucla.edu/">Computational
Vision and Learning Lab (CVL)</a> -
<strong><em>UCLA</em></strong>.</p></li>
<li><p><a href="http://web.cs.ucla.edu/~guyvdb/">Guy Van den Broeck</a>
- <strong><em>Department of Computer Science, UCLA</em></strong>, <a
href="http://starai.cs.ucla.edu/#">StarAI Lab</a> -
<strong><em>UCLA</em></strong>.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="uc-berkeley">UC Berkeley</h3>
<ul>
<li><p><a href="https://people.eecs.berkeley.edu/~anca/index.html">Anca
Dragan</a> - <strong><em>Department of Electrical Engineering and
Computer Science, UC Berkeley</em></strong>, <a
href="http://interact.berkeley.edu/">Interactive Autonomy and
Collaborative Technologies Laboratory (InterACT)</a> - <strong><em>UC
Berkeley</em></strong>.</p></li>
<li><p><a href="https://psychology.berkeley.edu/people/fei-xu">Fei
Xu</a> - <strong><em>Department of Psychology, UC
Berkeley</em></strong>, <a
href="https://babylab5.wixsite.com/bell">Berkeley Early Learning Lab (Xu
Lab)</a> - <strong><em>UC Berkeley</em></strong>.</p></li>
<li><p><a href="http://alisongopnik.com/">Alison Gopnik</a> -
<strong><em>Department of Psychology, UC Berkeley</em></strong>, <a
href="http://www.gopniklab.berkeley.edu/">Cognitive Development &amp;
Learning Lab (Gopnik Lab)</a> - <strong><em>UC
Berkeley</em></strong>.</p></li>
<li><p><a href="http://colala.berkeley.edu/people/piantadosi/">Steve
Piantadosi</a> - <strong><em>Department of Psychology, UC
Berkeley</em></strong>, <a href="http://colala.berkeley.edu/">The
computation and language lab (colala)</a> - <strong><em>UC
Berkeley</em></strong>.</p></li>
<li><p><a href="http://www.celestekidd.com/">Celeste Kidd</a> -
<strong><em>Department of Psychology, UC Berkeley</em></strong>, <a
href="https://www.kiddlab.com/">Kidd Lab</a> - <strong><em>UC
Berkeley</em></strong>.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="bnu">BNU</h3>
<ul>
<li><a
href="https://brain.bnu.edu.cn/English/Faculty/CurrentFaculty/Bzz/a552402e529a4f27b979378abd42c10e.htm">Yanchao
Bi</a> - <strong><em>IDG/McGovern Institute for Brain Research and the
State Key Laboratory of Cognitive Neuroscience and Learning, Beijing
Normal University (BNU)</em></strong>, <a
href="http://bilab.bnu.edu.cn/">Yanchao Bi’s Concept Lab (Bi Lab)</a> -
<strong><em>BNU</em></strong>.</li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="pku">PKU</h3>
<ul>
<li><p><a href="https://zhusongchun.net/">Song-Chun Zhu</a> -
<strong><em>School of AI and Institute for AI, Peking University
(PKU)</em></strong>.</p></li>
<li><p><a href="https://yzhu.io/">Yixin Zhu</a> - <strong><em>School of
AI and Institute for AI, Peking University (PKU)</em></strong>, <a
href="https://pku.ai/">Cognitive Reasoning Lab (CoRe Lab)</a> -
<strong><em>PKU</em></strong>.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="ucsd">UCSD</h3>
<ul>
<li><p><a href="https://pages.ucsd.edu/~ztu/">Zhuowen Tu</a> -
<strong><em>Department of Computer Science, UCSD</em></strong>, <a
href="https://pages.ucsd.edu/~ztu/Group.htm">Machine Learning,
Perception, and Cognition Lab (mlPC)</a> -
<strong><em>UCSD</em></strong>.</p></li>
<li><p><a
href="https://psychology.ucsd.edu/people/profiles/evul.html">Ed Vul</a>
- <strong><em>Department of Psychology, UCSD</em></strong>, <a
href="http://www.evullab.org/index.html">Computational Cognition Lab</a>
- <strong><em>UCSD</em></strong>.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="nyu">NYU</h3>
<ul>
<li><p><a href="https://cs.nyu.edu/~davise/">Ernest Davis</a> -
<strong><em>Department of Computer Science, Courant Institute of
Mathematical Sciences, NYU</em></strong>.</p></li>
<li><p><a href="http://garymarcus.com/index.html">Gary Marcus</a> -
<strong><em>Department of Psychology, NYU</em></strong>.</p></li>
<li><p><a href="https://cims.nyu.edu/~brenden/">Brenden Lake</a> -
<strong><em>Department of Psychology, NYU</em></strong>, <a
href="https://lake-lab.github.io/">Human &amp; Machine Learning Lab
(Lake Lab)</a> - <strong><em>NYU</em></strong>.</p></li>
<li><p><a href="https://as.nyu.edu/faculty/todd-gureckis.html">Todd
Gureckis</a> - <strong><em>Department of Psychology, NYU</em></strong>,
<a href="http://gureckislab.org/">Computation &amp; Cognition Lab</a> -
<strong><em>NYU</em></strong>.</p></li>
<li><p><a href="http://www.cns.nyu.edu/malab/people.html">Wei Ji Ma</a>
- <strong><em>Department of Psychology, Center for Neural Science,
NYU</em></strong>, <a href="http://www.cns.nyu.edu/malab/">Wei Ji Ma
Lab</a> - <strong><em>NYU</em></strong>.</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="jhu">JHU</h3>
<ul>
<li><a href="https://perception.jhu.edu/chaz/">Chaz Firestone</a> -
<strong><em>Department of Psychological and Brain Sciences, Johns
Hopkins University (JHU)</em></strong>, <a
href="https://perception.jhu.edu/">Hopkins Perception &amp; Mind Lab</a>
- <strong><em>JHU</em></strong>.</li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="sit">SIT</h3>
<ul>
<li><a href="https://markkho.github.io/">Mark Ho</a> -
<strong><em>Department of Computer Science, Stevens Institute of
Technology (SIT)</em></strong>, <a
href="https://codec-lab.github.io/">Computation and Decision-Making
Lab</a> - <strong><em>SIT</em></strong>.</li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h2 id="people-book">People &amp; Book</h2>
<h3 id="john-hopcroft">John Hopcroft</h3>
<p>Theoretical computer scientist.</p>
<ul>
<li><p><a
href="http://elib.vku.udn.vn/bitstream/123456789/2543/1/2007.%20Introduction%20to%20Automata%20Theory%2C%20Languages%2C%20and%20Computations%20%283rd%20edition%29.pdf">Introduction
to Automata Theory, Languages, and Computation</a> -
<strong><em>Pearson</em></strong>, 2007. [<a
href="https://scholar.google.com/scholar?cluster=326269839585842480">All
Versions</a>].</p></li>
<li><p><a
href="http://www.cs.cornell.edu/jeh/book%20no%20so;utions%20March%202019.pdf">Foundations
of Data Science</a> - <strong><em>Cambridge University
Press</em></strong>. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=1802704438630899850">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="ulf-grenander">Ulf Grenander</h3>
<p>Applied mathematician, the founder of General Pattern Theory.</p>
<ul>
<li><p><a
href="https://www.dam.brown.edu/ptg/REPORTS/calculustext.PDF">A Calculus
of Ideas: A Mathematical Study of Thinking</a> - <strong><em>World
Scientific Publishing Company</em></strong>, 2012. [<a
href="https://scholar.google.com/scholar?cluster=12182416000849265255&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://global.oup.com/academic/product/general-pattern-theory-9780198536710?cc=lt&amp;lang=de#">General
Pattern Theory: A Mathematical Study of Regular Structures</a> -
<strong><em>Oxford University Press</em></strong>, 1993. [<a
href="https://scholar.google.com/scholar?hl=en&amp;as_sdt=0%2C5&amp;q=General+Pattern+Theory&amp;btnG=">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="david-marr">David Marr</h3>
<p>Computational Cognitive Neuroscientist, the establisher of the Levels
of Analysis.</p>
<ul>
<li><a href="https://usa1lib.org/book/1223444/8e5ca8">Vision: A
Computational Investigation into the Human Representation and Processing
of Visual Information</a> - <strong><em>MIT Press</em></strong>, 1982.
[<a
href="https://scholar.google.com/scholar?cluster=14386368570811483142&amp;hl=en&amp;as_sdt=0,44">All
Versions</a>].</li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="michael-tomasello">Michael Tomasello</h3>
<p>Cognitive scientist, set up the foundations of studying human
communications.</p>
<ul>
<li><p><a href="https://1lib.net/book/541274/39859f">Origins of human
communication</a> - <strong><em>MIT Press</em></strong>, 2010. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=2553369883266458474">All
Versions</a>].</p></li>
<li><p><a
href="https://hk1lib.org/book/541275/1452f8?id=541275&amp;secret=1452f8">The
cultural origins of human cognition</a> - <strong><em>Havard University
Press</em></strong>, 2000. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=5000469061641945144">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="judea-pearl">Judea Pearl</h3>
<p>Applied mathematician, proposed causal intervention on siamese
bayesian networks.</p>
<ul>
<li><p><a href="http://bayes.cs.ucla.edu/WHY/">The Book of Why: The New
Science of Cause and Effect</a> - <strong><em>Basic Books</em></strong>,
2018. [<a
href="https://scholar.google.com/scholar?cluster=2505901292485349932&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a
href="https://hk1lib.org/book/2780725/2ec8f1?id=2780725&amp;secret=2ec8f1">Causality:
Models, Reasoning and Inference</a> - <strong><em>Cambridge University
Press</em></strong>, 2009. [<a
href="https://scholar.google.com/scholar?cluster=10996260119229499611&amp;hl=en&amp;as_sdt=0,5&amp;as_vis=1">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="susan-carey">Susan Carey</h3>
<p>Developmental psychologist, proposed <em>object</em> as a core
knowledge of human intelligence.</p>
<ul>
<li><p><a
href="https://hk1lib.org/book/844457/42178f?id=844457&amp;secret=42178f">The
Origin of Concepts</a> - <strong><em>Oxford University
Press</em></strong>, 2009. [<a
href="https://scholar.google.com/scholar?cluster=11493102398422813821&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://hk1lib.org/book/3659332/11fa44">Conceptual
Change in Childhood</a> - <strong><em>MIT Press</em></strong>, 1985. [<a
href="https://scholar.google.com/scholar?hl=en&amp;as_sdt=0%2C5&amp;q=conceptual+change+in+childhood+susan+carey&amp;btnG=">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="daniel-kahneman">Daniel Kahneman</h3>
<p>Computational cognitive scientist and Economist, set up the
foundations for Decision Theory.</p>
<ul>
<li><a
href="https://hk1lib.org/book/2181569/f5e85a?id=2181569&amp;secret=f5e85a">Thinking,
fast and slow</a> - <strong><em>Farrar Straus Giroux</em></strong>,
2011. [<a
href="https://scholar.google.com/scholar?oi=bibs&amp;hl=en&amp;cluster=3255681708785115121">All
Versions</a>].</li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h3 id="karl-popper">Karl Popper</h3>
<p>Scientific philosophor, the founder of scientific verification
theories.</p>
<ul>
<li><p><a href="https://hk1lib.org/book/511214/299596">The logic of
scientific discovery</a> - <strong><em>Routledge</em></strong>, 2005.
[<a
href="https://scholar.google.com/scholar?cluster=5836864564733788424&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
<li><p><a href="https://hk1lib.org/book/2773070/c48f60">All Life is
Problem Solving</a> - <strong><em>Routledge</em></strong>, 2001. [<a
href="https://scholar.google.com/scholar?cluster=9799073870888093350&amp;hl=en&amp;as_sdt=0,5">All
Versions</a>].</p></li>
</ul>
<p>*<a href="#c">Back to Top</a></p>
<h2 id="about">About</h2>
<p>The initiator of this repo has been struggling to taxonomize related
topics, since there are so many perspectives to follow, such as
task-oriented, technique-oriented, and metaphysics-oriented. Finally he
decided to focus on the perspective of <strong><em>The Sciences of
Intelligence</em></strong>—each topic describes a phenomenon of
intelligence, or an intelligent behavior—they show the objectives of
reverse-engineering human intelligence for computational methods. These
topics are never restricted to specific technical methods or tasks, but
are trying to organize the nature of intelligence—from both <em>the
software perspective</em> and <em>the hardware perspective</em>.</p>
<p>Obviously, this reading list is far from covering the every aspect of
AGI and CoCoSci. Since the list is a by-product of the literature
reviews when the initiator is working on Abduction and Bayesian
modeling, other topics are also collected with biases, more or less.
Abduction may be the way humans explain the world with the known, and
discover the unknown, requiring much more investigations into its
computational basis, cognitive underpinnings, and applications to AI.
Please feel free to reach out!</p>
<p>*<a href="#c">Back to Top</a></p>
<p><a
href="https://github.com/YuzheSHI/awesome-agi-cocosci">agicocosci.md
Github</a></p>
Reference in New Issue View Git Blame Copy Permalink