Human few-shot learning of compositional instructions

Brenden M. Lake, Tal Linzen, Marco Baroni

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

People learn in fast and flexible ways that have not been emulated by machines. Once a person learns a new verb “dax,” he or she can effortlessly understand how to “dax twice,” “walk and dax,” or “dax vigorously.” There have been striking recent improvements in machine learning for natural language processing, yet the best algorithms require vast amounts of experience and struggle to generalize new concepts in compositional ways. To better understand these distinctively human abilities, we study the compositional skills of people through language-like instruction learning tasks. Our results show that people can learn and use novel functional concepts from very few examples (few-shot learning), successfully applying familiar functions to novel inputs. People can also compose concepts in complex ways that go beyond the provided demonstrations. Two additional experiments examined the assumptions and inductive biases that people make when solving these tasks, revealing three biases: mutual exclusivity, one-to-one mappings, and iconic concatenation. We discuss the implications for cognitive modeling and the potential for building machines with more human-like language learning capabilities.

Original languageEnglish (US)
Title of host publicationProceedings of the 41st Annual Meeting of the Cognitive Science Society
Subtitle of host publicationCreativity + Cognition + Computation, CogSci 2019
PublisherThe Cognitive Science Society
Pages611-617
Number of pages7
ISBN (Electronic)0991196775, 9780991196777
StatePublished - 2019
Event41st Annual Meeting of the Cognitive Science Society: Creativity + Cognition + Computation, CogSci 2019 - Montreal, Canada
Duration: Jul 24 2019Jul 27 2019

Publication series

NameProceedings of the 41st Annual Meeting of the Cognitive Science Society: Creativity + Cognition + Computation, CogSci 2019

Conference

Conference41st Annual Meeting of the Cognitive Science Society: Creativity + Cognition + Computation, CogSci 2019
Country/TerritoryCanada
CityMontreal
Period7/24/197/27/19

Keywords

  • compositionality
  • concept learning
  • neural networks
  • word learning

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Science Applications
  • Human-Computer Interaction
  • Cognitive Neuroscience

Fingerprint

Dive into the research topics of 'Human few-shot learning of compositional instructions'. Together they form a unique fingerprint.

Cite this