What underlies rapid learning and systematic generalization in humans

Despite the groundbreaking successes of neural networks, contemporary models require extensive training with massive datasets and exhibit poor out-of-sample generalization. One proposed solution is to build systematicity and domain-specific constraints into the model, echoing the tenets of classical, symbolic cognitive architectures. In this paper, we consider the limitations of this approach by examining human adults’ ability to learn an abstract reasoning task from a brief instructional tutorial and explanatory feedback for incorrect responses, demonstrating that human learning dynamics and ability to generalize outside the range of the training examples differ drastically from those of a representative neural network model, and that the model is brittle to changes in features not anticipated by its authors. We present further evidence from human data that the ability to consistently solve the puzzles was associated with education, particularly basic mathematics education, and with the ability to provide a reliably identifiable, valid description of the strategy used. We propose that rapid learning and systematic generalization in humans may depend on a gradual, experience-dependent process of learningto-learn using instructions and explanations to guide the construction of explicit abstract rules that support generalizable inferences.

[1]  Hinrich Schütze,et al.  Placing language in an integrated understanding system: Next steps toward human-level performance in neural language models , 2020, Proceedings of the National Academy of Sciences.

[2]  Wojciech M. Czarnecki,et al.  Grandmaster level in StarCraft II using multi-agent reinforcement learning , 2019, Nature.

[3]  P C Wason,et al.  Reasoning about a Rule , 1968, The Quarterly journal of experimental psychology.

[4]  Zeb Kurth-Nelson,et al.  Learning to reinforcement learn , 2016, CogSci.

[5]  John Field,et al.  Language and the mind , 1968 .

[6]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[7]  Shane Legg,et al.  Human-level control through deep reinforcement learning , 2015, Nature.

[8]  Jessica B. Hamrick,et al.  psiTurk: An open-source framework for conducting replicable behavioral experiments online , 2016, Behavior research methods.

[9]  Demis Hassabis,et al.  Mastering the game of Go with deep neural networks and tree search , 2016, Nature.

[10]  Joshua B. Tenenbaum,et al.  Human Learning in Atari , 2017, AAAI Spring Symposia.

[11]  Bertram Felgenhauer,et al.  Mathematics of Sudoku I , 2006 .

[12]  Matthew Crosby,et al.  Association for the Advancement of Artificial Intelligence , 2014 .

[13]  Amos Storkey,et al.  Meta-Learning in Neural Networks: A Survey , 2020, IEEE transactions on pattern analysis and machine intelligence.

[14]  A NEWELL,et al.  Computer simulation of human thinking. , 1961, Science.

[15]  P. Johnson-Laird Mental models and deduction , 2001, Trends in Cognitive Sciences.

[16]  G. Marcus The Algebraic Mind: Integrating Connectionism and Cognitive Science , 2001 .

[17]  Jürgen Schmidhuber,et al.  Multi-column deep neural networks for image classification , 2012, 2012 IEEE Conference on Computer Vision and Pattern Recognition.

[18]  Allen Newell,et al.  Report on a general problem-solving program , 1959, IFIP Congress.

[19]  R. Mooney,et al.  Schema acquisition from a single example , 1992 .

[20]  Joshua B. Tenenbaum,et al.  Learning Structured Generative Concepts , 2010 .

[21]  Felix Hill,et al.  Imitating Interactive Intelligence , 2020, ArXiv.

[22]  M. Cole Cognitive Consequences of Formal and Informal Education , 2008 .

[23]  Sergey Levine,et al.  Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks , 2017, ICML.

[24]  M. Cole,et al.  The cultural context of learning and thinking. An Exploration in Experimental Anthropology. , 1971 .

[25]  Ole Winther,et al.  Recurrent Relational Networks , 2017, NeurIPS.

[26]  D. Laplane Thought and language. , 1992, Behavioural neurology.

[27]  R. Mooney,et al.  Explanation-Based Learning: An Alternative View , 1986, Machine Learning.

[28]  Marco Baroni,et al.  Generalization without Systematicity: On the Compositional Skills of Sequence-to-Sequence Recurrent Networks , 2017, ICML.

[29]  Joshua B. Tenenbaum,et al.  Building machines that learn and think like people , 2016, Behavioral and Brain Sciences.

[30]  James L. McClelland,et al.  Parallel Distributed Processing at 25: Further Explorations in the Microstructure of Cognition , 2014, Cogn. Sci..

[31]  Felix Hill,et al.  Learning to Make Analogies by Contrasting Abstract Relational Structure , 2019, ICLR.

[32]  Zenon W. Pylyshyn,et al.  Connectionism and cognitive architecture: A critical analysis , 1988, Cognition.

[33]  E. Spelke,et al.  Language and Conceptual Development series Core systems of number , 2004 .

[34]  Todd M. Gureckis,et al.  psiTurk: An open platform for science on Amazon Mechanical Turk , 2021 .

[35]  Joshua B. Tenenbaum,et al.  Human-level concept learning through probabilistic program induction , 2015, Science.

[36]  Paul-Christian Bürkner,et al.  brms: An R Package for Bayesian Multilevel Models Using Stan , 2017 .

[37]  James L. McClelland,et al.  An interactive activation model of context effects in letter perception: I. An account of basic findings. , 1981 .

[38]  Sergio Gomez Colmenarejo,et al.  Hybrid computing using a neural network with dynamic external memory , 2016, Nature.

[39]  Pushmeet Kohli,et al.  Analysing Mathematical Reasoning Abilities of Neural Models , 2019, ICLR.

[40]  M. Chi,et al.  The ICAP Framework: Linking Cognitive Engagement to Active Learning Outcomes , 2014 .

[41]  S. Pinker,et al.  On language and connectionism: Analysis of a parallel distributed processing model of language acquisition , 1988, Cognition.

[42]  Philipp Slusallek,et al.  Introduction to real-time ray tracing , 2005, SIGGRAPH Courses.

[43]  Mark Chen,et al.  Language Models are Few-Shot Learners , 2020, NeurIPS.

[44]  Razvan Pascanu,et al.  A simple neural network module for relational reasoning , 2017, NIPS.

[45]  Lukasz Kaiser,et al.  Attention is All you Need , 2017, NIPS.

[46]  Ivan Vankov,et al.  Training neural networks to encode symbols enables combinatorial generalization , 2019, Philosophical Transactions of the Royal Society B.