Open-ended Scientific Discovery via Bayesian Surprise
Dhruv Agarwal Bodhisattwa Prasad Majumder, Reece Adamson Megha Chakravorty Satvika Reddy Gavireddy, Aditya Parashar Harshit Surana Bhavana Dalvi Mishra, Andrew McCallum Ashish Sabharwal Peter Clark · 2025
Read Open-ended Scientific Discovery via Bayesian Surprise by Dhruv Agarwal Bodhisattwa Prasad Majumder, Reece Adamson Megha Chakravorty Satvika Reddy Gavireddy, Aditya Parashar Harshit Surana Bhavana Dalvi Mishra, Andrew McCallum Ashish Sabharwal Peter Clark (2025) 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-2507-00310.txt.
Description
Dhruv Agarwal and co-authors present a system that chooses its own science questions in a 2025 paper. Most robot science tools need a human to set the research question first. Open-ended tools exist, but they pick ideas by variety or by vague guesses at what humans find interesting. Both paths can waste effort.
Their system, AutoDiscovery, uses Bayesian surprise as its guide. It measures how much a result shifts the model's beliefs about an idea. A tree search then picks the next test under a fixed budget. Across 21 real data sets, the system found more discoveries than rivals, the authors report.
Catalog record
| CREATOR | Dhruv Agarwal Bodhisattwa Prasad Majumder, Reece Adamson Megha Chakravorty Satvika Reddy Gavireddy, Aditya Parashar Harshit Surana Bhavana Dalvi Mishra, Andrew McCallum Ashish Sabharwal Peter Clark |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 88 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-00310