Guy Davidson
Guy Davidson

Research Scientist

FAIR, Meta Superintelligence Labs

About Me

I’m a cognitive scientist and machine learning researcher working as a Research Scientist at FAIR under Meta Superintelligence Labs, where I resesarch topics around evaluating (and improving) theory of mind and other mental modelling capabilities in large language models. Prior to this I did my PhD at the NYU Center for Data Science, advised by Brenden Lake and Todd Gureckis. My dissertation, titled “Goals as Reward Producing Programs”, offers theoretical, empirical, and computational advances in the study of goals: how do we represent, reason about, and come up with them? In parallel, I spent the last year of my Phd working on mechanistic interpretability as a visiting researcher at Meta FAIR with Adina Williams.

In my non-academic life, I live with my wife Sarah and our dog Lila, and spend time making homemade hot sauces, playing board games, and playing ultiamte frisbee (poorly) and lifting weights (slightly less poorly).

Download CV
Interests
  • Goal representation and generation
  • Human-like goals for agents
  • Compuational cognitive science
  • Goal/intent inference in LLMs
Education

  • PhD in Data Science

    New York University

  • MPhil in Data Science

    New York University

  • BSc in Computational Sciences

    Minerva University

Projects
Do different prompting methods yield a common task representation in language models?
Goals as Reward-Producing Programs
Infant Relation Categorization
Playful Goal Modeling
Simple Object Representations
Publications
Guy Davidson, Todd M. Gureckis, Brenden M. Lake, Adina Williams (2025). Do different prompting methods yield a common task representation in language models?. Advances in Neural Information Processing Systems 38 (NeurIPS 2025).
Cite URL
Yueh-Han Chen, Guy Davidson, Brenden M. Lake (2025). SAGE-Eval: Evaluating LLMs for Systematic Generalizations of Safety Facts. Advances in Neural Information Processing Systems 38 (NeurIPS 2025).
Cite URL
Guy Davidson, Graham Todd, Cedric Colas, Junyi Chu, Julian Togelius, Joshua B. Tenenbaum, Todd M. Gureckis, Brenden M. Lake (2025). Goal Inference using Reward-Producing Programs in a Novel Physics Environment. Proceedings of the 47th Annual Meeting of the Cognitive Science Society, CogSci 2025.
PDF Cite
Graham Todd, Junyi Chu, Guy Davidson, Weijia Xu (2025). Novel Goal Creation and Evaluation in Open-Ended Games. Forthcoming in the Proceedings of the 47th Annual Meeting of the Cognitive Science Society, CogSci 2025.
PDF Cite
Guy Davidson, Graham Todd, Julian Togelius, Todd M. Gureckis, Brenden M. Lake (2025). Goals as Reward-Producing Programs. Nature Machine Intelligence.
Cite URL
Xiang Pan, Jason Phang, Guy Davidson, Ethan Perez (2025). Generate-Feedback-Refine: How Much Does Model Quality in Each Role Matter?. Deep Learning 4 Code @ ICLR 2025.
Cite URL
Guy Davidson, Todd M. Gureckis (2025). Position: Human-like reinforcement learning ıs facilitated by structured and expressive goals.
PDF Cite
Guy Davidson, Todd M. Gureckis (2024). Toward Complex and Structured Goals in Reinforcement Learning. Finding the Frame @ RLC 2024.
Cite URL
Guy Davidson, A. Emin Orhan, Brenden M. Lake (2024). Spatial relation categorization in infants and deep neural networks. Cognition.
Cite DOI URL
Sugandha Sharma, Guy Davidson, Khimya Khetarpal, Anssi Kanervisto, Udit Arora, Katja Hofmann, Ida Momennejad (2024). Toward Human-AI Alignment in Large-Scale Multi-Player Games. Wordplay @ ACL 2024, Association for Computational Linguistics.
PDF Cite arXiv
Guy Davidson, Graham Todd, Todd M. Gureckis, Julian Togelius, Brenden M. Lake (2023). Generating Human-Like Goals by Synthesizing Reward-Producing Programs. Intrinsically Motivated Open-Ended Learning @ NeurIPS 2023.
PDF Cite
Guy Davidson, Todd M. Gureckis, Brenden M. Lake (2022). Creativity, Compositionality, and Common Sense in Human Goal Generation. Proceedings of the 44th Annual Meeting of the Cognitive Science Society, CogSci 2022.
Cite URL
Daniel Bennett, Guy Davidson, Yael Niv (2022). A model of mood as integrated advantage. Psychological Review.
Cite DOI URL
Guy Davidson, Brenden M. Lake (2021). Examining Infant Relation Categorization Through Deep Neural Networks. Proceedings of the 43rd Annual Meeting of the Cognitive Science Society, CogSci 2021.
Cite URL
Guy Davidson, Brenden M. Lake (2020). Investigating Simple Object Representations in Model-Free Deep Reinforcement Learning. Proceedings of the 42nd Annual Meeting of the Cognitive Science Society, CogSci 2020.
PDF Cite arXiv
Guy Davidson, Brenden M. Lake (2020). Systematically Comparing Neural Network Architectures in Relation Learning. Object-Oriented Learning (OOL): Perception, Representation, and Reasoning @ ICML 2020.
Cite URL
Guy Davidson, Michael C. Mozer (2020). Sequential mastery of multiple visual tasks: Networks naturally learn to learn and forget to forget . The IEEE Conference on Computer Vision and Pattern Recognition (CVPR).
Cite arXiv URL
Guy Davidson, Angela Radulescu, Yael Niv (2019). Contrasting the effects of prospective attention and retrospective decay in representation learning. The 4th Multidisciplinary Conference on Reinforcement Learning and Decision Making.
PDF Cite
Daniel Bennett, Guy Davidson, Yael Niv (2019). Momentum and mood in policy-gradient reinforcement learning. The 4th Multidisciplinary Conference on Reinforcement Learning and Decision Making.
PDF Cite