auto-psych : Automating the science of mind using agent-driven theory discovery and experimentation
Ben Prystawski, Kushin Mukherjee, Daniel Wurgaft, Linas Nasvytis, Michael Y. Li, Noah D. Goodman, Michael C. Frank · 2026
Read auto-psych : Automating the science of mind using agent-driven theory discovery and experimentation by Ben Prystawski, Kushin Mukherjee, Daniel Wurgaft, Linas Nasvytis, Michael Y. Li, Noah D. Goodman, Michael C. Frank (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-26460.txt.
Description
Ben Prystawski and six colleagues present auto-psych, a system that automates theory building in cognitive psychology. Language model agents can already propose hypotheses and analyse data. Collecting human data was the bottleneck. Psychology is well placed for automation, since its theories are often written as code. Crowdsourcing platforms also allow experiments to run online at scale.
Auto-psych runs two nested loops. The inner loop proposes probabilistic cognitive models, fits them, and critiques them. Experiments are designed by the outer loop, launched online, and analysed there. The test case is a classic puzzle: which coin-flip sequences people judge as more random. The system recovered known true theories from synthetic data reliably. The nested structure proved essential to that success. In three rounds of human experiments, it found theories fitting the data better than published ones.
Catalog record
| CREATOR | Ben Prystawski, Kushin Mukherjee, Daniel Wurgaft, Linas Nasvytis, Michael Y. Li, Noah D. Goodman, Michael C. Frank |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 85 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.26460