Using games to understand the mind

K Allen, F Brändle, M Botvinick, JE Fan… - Nature Human …, 2024 - nature.com
Board, card or video games have been played by virtually every individual in the world.
Games are popular because they are intuitive and fun. These distinctive qualities of games …

Using natural language and program abstractions to instill human inductive biases in machines

S Kumar, CG Correa, I Dasgupta… - Advances in …, 2022 - proceedings.neurips.cc
Strong inductive biases give humans the ability to quickly learn to perform a variety of tasks.
Although meta-learning is a method to endow neural networks with useful inductive biases …

Abstract visual reasoning with tangram shapes

A Ji, N Kojima, N Rush, A Suhr, WK Vong… - arXiv preprint arXiv …, 2022 - arxiv.org
We introduce KiloGram, a resource for studying abstract visual reasoning in humans and
machines. Drawing on the history of tangram puzzles as stimuli in cognitive science, we …

From partners to populations: A hierarchical Bayesian account of coordination and convention.

RD Hawkins, M Franke, MC Frank… - Psychological …, 2023 - psycnet.apa.org
Languages are powerful solutions to coordination problems: They provide stable, shared
expectations about how the words we say correspond to the beliefs and intentions in our …

Communicating natural programs to humans and machines

S Acquaviva, Y Pu, M Kryven… - Advances in …, 2022 - proceedings.neurips.cc
Abstract The Abstraction and Reasoning Corpus (ARC) is a set of procedural tasks that tests
an agent's ability to flexibly solve novel problems. While most ARC tasks are easy for …

Reconciling truthfulness and relevance as epistemic and decision-theoretic utility.

TR Sumers, MK Ho, TL Griffiths… - Psychological Review, 2024 - psycnet.apa.org
People use language to influence others' beliefs and actions. Yet models of communication
have diverged along these lines, formalizing the speaker's objective in terms of either the …

DiffVL: scaling up soft body manipulation using vision-language driven differentiable physics

Z Huang, F Chen, Y Pu, C Lin… - Advances in Neural …, 2023 - proceedings.neurips.cc
Combining gradient-based trajectory optimization with differentiable physics simulation is an
efficient technique for solving soft-body manipulation problems. Using a well-crafted …

Learning with language-guided state abstractions

A Peng, I Sucholutsky, BZ Li, TR Sumers… - arXiv preprint arXiv …, 2024 - arxiv.org
We describe a framework for using natural language to design state abstractions for
imitation learning. Generalizable policy learning in high-dimensional observation spaces is …

Identifying concept libraries from language about object structure

C Wong, WP McCarthy, G Grand, Y Friedman… - arXiv preprint arXiv …, 2022 - arxiv.org
Our understanding of the visual world goes beyond naming objects, encompassing our
ability to parse objects into meaningful parts, attributes, and relations. In this work, we …

Disentangling abstraction from statistical pattern matching in human and machine learning

S Kumar, I Dasgupta, ND Daw, JD Cohen… - PLoS computational …, 2023 - journals.plos.org
The ability to acquire abstract knowledge is a hallmark of human intelligence and is believed
by many to be one of the core differences between humans and neural network models …