2021

Abstraction and Analogy-Making in Artificial Intelligence

Mitchell, Melanie

Understand

Conceptual abstraction and analogy-making are key abilities underlying humans' abilities to learn, reason, and robustly adapt their knowledge to new domains.

Despite of a long history of research on constructing AI systems with these abilities, no current AI system is anywhere close to a capability of forming humanlike abstractions or analogies.
This paper reviews the advantages and limitations of several approaches toward this goal, including symbolic methods, deep learning, and probabilistic program induction.
The paper concludes with several proposals for designing challenge tasks and evaluation measures in order to make quantifiable and generalizable progress in this area.

Built on

On the measure of intelligence
Original
F. Chollet · 1911
Earlier work this paper cites.
The Computer and the Brain
J. von Neumann · 1958
Earlier work this paper cites.
A program for the solution of a class of geometric-analogy intelligence-test questions
T. G. Evans · 1968
Earlier work this paper cites.
Pattern Recognition
M. M. Bongard · 1970
Earlier work this paper cites.
Gödel, Escher, Bach: an Eternal Golden Braid
D. R. Hofstadter · 1979
Earlier work this paper cites.
The Hearsay-II speech-understanding system: Integrating knowledge to resolve uncertainty
L. D. Erman, F. Hayes-Roth, V. R. Lesser, and D. R. Reddy · 1980
Earlier work this paper cites.
Darwin and the political economists: Divergence of character
S. S. Schweber · 1980
Earlier work this paper cites.
Learning and reasoning by analogy
P. H. Winston · 1980
Earlier work this paper cites.
Structure-mapping: A theoretical framework for analogy
D. Gentner · 1983
Earlier work this paper cites.
Analogical thinking and human intelligence
K. J. Holyoak · 1984
Earlier work this paper cites.
Analogies and roles in human and machine thinking
D. R. Hofstadter · 1985
Earlier work this paper cites.
The structure-mapping engine
B. Falkenhainer, K. D. Forbus, and D. Gentner · 1986
Earlier work this paper cites.
Attention, similarity, and the identification–categorization relationship
R. M Nosofsky · 1986
Earlier work this paper cites.
Analogical mapping by constraint satisfaction
K. J. Holyoak and P. Thagard · 1989
Earlier work this paper cites.
What one intelligence test measures: A theoretical account of the processing in the Raven progressive matrices test
P. A. Carpenter, M. A. Just, and P. Shell · 1990
Earlier work this paper cites.
Tensor product variable binding and the representation of symbolic structures in connectionist systems
P. Smolensky · 1990
Earlier work this paper cites.
High-level perception, representation, and analogy: A critique of artificial intelligence methodology
D. J. Chalmers, R. M. French, and D. R. Hofstadter · 1992
Earlier work this paper cites.
The reviewing of object files: Object-specific integration of information
D. Kahneman, A. Treisman, and B. J. Gibbs · 1992
Earlier work this paper cites.
How to build a baby: II. Conceptual primitives
J. M. Mandler · 1992
Earlier work this paper cites.
Analogy-Making as Perception: A Computer Model
M. Mitchell · 1993
Earlier work this paper cites.
Advances in Connectionist and Neural Computation Theory Vol. 2: Volume Two: Analogical Connections
J. A. Barnden and K. J. Holyoak · 1994
Earlier work this paper cites.
The Copycat project: A model of mental fluidity and analogy-making
D. R. Hofstadter and M. Mitchell · 1994
Earlier work this paper cites.
Fluid Concepts and Creative Analogies: Computer Models of the Fundamental Mechanisms of Thought
D. R. Hofstadter · 1995
Earlier work this paper cites.
LISA: A computational model of analogical inference and schema induction
J. E. Hummel and K. J. Holyoak · 1996
Earlier work this paper cites.
A concept learning algorithm with adaptive search
K. Saito and R. Nakano · 1996
Earlier work this paper cites.
Mappings in Thought and Language
G. Fauconnier · 1997
Earlier work this paper cites.
Conceptual integration in counterfactuals
M. Turner and G. Fauconnier · 1997
Earlier work this paper cites.
Analogy just looks like high level perception: Why a domain-general approach to analogical mapping is right
K. D. Forbus, D. Gentner, A. B. Markman, and R. W. Ferguson · 1998
Earlier work this paper cites.
Feature binding, attention and object perception
A. Treisman · 1998
Earlier work this paper cites.
Principles of categorization
E. Rosch · 1999
Earlier work this paper cites.
Analogy as the core of cognition
D. R. Hofstadter · 2001
Earlier work this paper cites.
Integrating memory and reasoning in analogy-making: The AMBR model
B. N. Kokinov and A. Petrov · 2001
Earlier work this paper cites.
The computational modeling of analogy-making
R. M. French · 2002
Earlier work this paper cites.
How conscious experience and working memory interact
B. J. Baars and S. Franklin · 2003
Earlier work this paper cites.
The theory theory as an alternative to the innateness hypothesis
A. Gopnik · 2003
Earlier work this paper cites.
Analogical learning of visual/conceptual relationships in sketches
K. D. Forbus, J. Usher, and E. Tomai · 2005
Earlier work this paper cites.
PHAEACO: A Cognitive Architecture Inspired by Bongard’s Problems
H. E. Foundalis · 2006
Earlier work this paper cites.
A Proposal for the Dartmouth Summer Research Project in Artificial Intelligence, 1955
J. McCarthy, M. L. Minsky, N. Rochester, and C. E. Shannon · 2006
Earlier work this paper cites.

Similar

A cognitive architecture that combines internal simulation with a global workspace
M. Shanahan · 2006
Cited alongside, same era.
Brain states: Top-down influences in sensory processing
C. D. Gilbert and M. Sigman · 2007
Cited alongside, same era.
Core knowledge
E. S. Spelke and K. D. Kinzler · 2007
Cited alongside, same era.
An integrated reasoning approach to moral decision making
M. Dehghani, K. D. Forbus, E. Tomai, and M. Klenk · 2008
Cited alongside, same era.
CogSketch
K. D. Forbus, A. Lovett, K. Lockwood, J. Wetzel, C. Matuk, B. Jee, and J. Usher · 2008
Cited alongside, same era.
The contemporary theory of metaphor
G. Lakoff · 2008
Cited alongside, same era.
Simulation, situated conceptualization, and prediction
L. W. Barsalou · 2009
Cited alongside, same era.
Musicat: A model of music perception and expectation
E. Nichols and D. Hofstadter · 2009
Cited alongside, same era.
Précis of The Origin of Concepts
S. Carey · 2011
Cited alongside, same era.
The neural theory of metaphor
G. Lakoff · 2012
Cited alongside, same era.
Surfaces and Essences
D. Hofstadter and E. Sander · 2013
Cited alongside, same era.
Linguistic regularities in continuous space word representations
T. Mikolov, W. T. Yih, and G. Zweig · 2013
Cited alongside, same era.

Then

Analogy and analogical reasoning
P. Bartha · 2019
Later among the works it cites.
Execution-guided neural program synthesis
X. Chen, C. Liu, and D. Song · 2019
Later among the works it cites.
Attention on abstract visual reasoning
Original
L. Hahne, T. Lüddecke, F. Wörgötter, and D. Kappel · 2019
Later among the works it cites.
Learning to make analogies by contrasting abstract relational structure
F. Hill, A. Santoro, D. G. T. Barrett, A. Morcos, and T. Lillicrap · 2019
Later among the works it cites.
Beyond imitation: Zero-shot task transfer on robots by learning concepts as cognitive programs
M. Lázaro-Gredilla, D. Lin, J. S. Guntupalli, and D. George · 2019
Later among the works it cites.
Seeing the meaning : Vision meets semantics in solving pictorial analogy problems
H. Lu, Q. Liu, N. Ichien, A. L. Yuille, and K. J. Holyoak · 2019
Later among the works it cites.
The neuro-symbolic concept learner: Interpreting scenes, words, and sentences from natural supervision
J. Mao, C. Gan, P. Kohli, J. B. Tenenbaum, and J. Wu · 2019
Later among the works it cites.
Artificial Intelligence: A Guide for Thinking Humans, Chapter 6
M. Mitchell · 2019
Later among the works it cites.
Detecting unseen visual relations using analogies
J. Peyre, J. Sivic, I. Laptev, and C. Schmid · 2019
Later among the works it cites.
SuperGLUE: A stickier benchmark for general-purpose language understanding systems
A. Wang, Y. Pruksachatkun, N. Nangia, J. Singh, A.and Michael, F. Hill, O. Levy, and S. R. Bowman · 2019
Later among the works it cites.
RAVEN: A dataset for relational and analogical visual reasoning
C. Zhang, F. Gao, B. Jia, Y. Zhu, and S.-C. Zhu · 2019
Later among the works it cites.
Abstract reasoning with distracting features
K. Zheng, Z.-J. Zha, and W. Wei · 2019
Later among the works it cites.
We have so much in common: Modeling semantic relational set abstractions in videos
Original
A. Andonian, C. Fosco, M. Monfort, A. Lee, R. Feris, C. Vondrick, and A. Oliva · 2020
Later among the works it cites.
Personal Communication, 2020
F. Chollet · 2020
Later among the works it cites.
Neural analogical matching
Original
M. Crouse, C. Nakos, I. Abdelaziz, and K. D. Forbus · 2020
Later among the works it cites.
Relation learning in a neurocomputational architecture supports cross-domain transfer
L. A. A. Doumas, G. Puebla, A. E. Martin, and J. E. Hummel · 2020
Later among the works it cites.
DreamCoder: Growing generalizable, interpretable knowledge with wake-sleep Bayesian program learning
Original
K. Ellis, C. Wong, M. Nye, M. Sablé-Meyer, L. Cary, L. Morales, L. Hewitt, A. Solar-Lezama, and J. B. Tenenbaum · 2020
Later among the works it cites.
Five points to check when comparing visual perception in humans and machines
Original
C. M. Funke, J. Borowski, K. Stosio, W. Brendel, T. S.A. Wallis, and M. Bethge · 2020
Later among the works it cites.
Shortcut learning in deep neural networks
R. Geirhos, J. H. Jacobsen, C. Michaelis, R. Zemel, W. Brendel, M. Bethge, and F. A. Wichmann · 2020
Later among the works it cites.
Environmental drivers of systematicity and generalization in a situated agent
F. Hill, A. Lampinen, R. Schneider, S. Clark, M. Botvinick, J. L. McClelland, and A. Santoro · 2020
Later among the works it cites.
Solving Raven’s Progressive Matrices with Multi-Layer Relation Networks
M. Jahrens and T. Martinetz · 2020
Later among the works it cites.
Transforming task representations to perform novel tasks
A. K. Lampinen and J. L. McClelland · 2020
Later among the works it cites.
How can we accelerate progress towards human-like linguistic generalization?
T. Linzen · 2020
Later among the works it cites.
Right for the wrong reasons: Diagnosing syntactic heuristics in natural language inference
T. McCoy, E. Pavlick, and T. Linzen · 2020
Later among the works it cites.
Bongard-LOGO: A new benchmark for human-level concept learning and reasoning
Original
W. Nie, Z. Yu, A. B. Mao, L.and Patel, Y. Zhu, and A. Anandkumar · 2020
Later among the works it cites.
Joe Biden positions himself as ‘bridge’ to next generation at Michigan rally
T. Perkins · 2020
Later among the works it cites.
The child as hacker
J. S. Rule, J. B. Tenenbaum, and S. T. Piantadosi · 2020
Later among the works it cites.
Winogrande: An adversarial Winograd schema challenge at scale
K. Sakaguchi, R. Le Bras, C. Bhagavatula, and Y. Choi · 2020
Later among the works it cites.
Meta-learning in natural and artificial intelligence
Original
J. X. Wang · 2020
Later among the works it cites.
Learning representations that support extrapolation
Original
T. W. Webb, Z. Dulberg, S. M. Frankland, A. A. Petrov, R. C. O’Reilly, and J. D. Cohen · 2020
Later among the works it cites.
The Scattering Compositional Learner: Discovering Objects, Attributes, Relationships in Analogical Reasoning
Original
Y. Wu, H. Dong, R. Grosse, and J. Ba · 2020
Later among the works it cites.
Solving Raven’s progressive matrices with neural networks
Original
T. Zhuo and . Kankanhalli · 2020
Later among the works it cites.
Making sense of sensory input
R. Evans, J. Hernández-Orallo, J. Welbl, P. Kohli, and M. Sergot · 2021
Closest in time.
Learning task-general representations with generative neuro-symbolic modeling
R. Feinman and B. M. Lake · 2021
Closest in time.
Grounded language learning fast and slow
F. Hill, O. Tieleman, T. von Glehn, N. Wong, H. Merzic, and S. Clark · 2021
Closest in time.
Stratified rule-aware network for ab stract visual reasoning
S. Hu, Y. Ma, X. Liu, Y. Wei, and S. Bai · 2021
Closest in time.
Emergent symbols through binding in external memory
T. W. Webb, I. Sinha, and J. D. Cohen · 2021
Closest in time.
Abstract spatial-temporal reasoning via probabilistic abduction and execution
Original
C. Zhang, B. Jia, S. Zhu, and Y. Zhu · 2021
Closest in time.

Beyond the bibliography

alphaXiv searches the wider corpus for related work and actual follow-ups.

Open on alphaXiv

alphaXiv is searching for related work…

Reading the bibliography…