Stargazer : A Scalable Model-Fitting Benchmark Environment for AI Agents under Astrophysical Constraints
Xinge Liu, Terry Jingchen Zhang, Bernhard Schölkopf, Zhijing Jin, Kristen Menou · 2026
Read Stargazer : A Scalable Model-Fitting Benchmark Environment for AI Agents under Astrophysical Constraints by Xinge Liu, Terry Jingchen Zhang, Bernhard Schölkopf, Zhijing Jin, Kristen Menou (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-2604-15664.txt.
Description
Liu and four colleagues introduce Stargazer, a scalable benchmark environment for AI agents doing physics-grounded model fitting on radial-velocity time series. Its 120 tasks across three tiers, including 20 real archival cases, range from high-signal single planets to complex multi-planet low-signal analyses. Eight frontier agents often fit data well yet miss true physical parameters, and extra test-time compute helps only marginally, with token bloat signaling failure loops.
Catalog record
| CREATOR | Xinge Liu, Terry Jingchen Zhang, Bernhard Schölkopf, Zhijing Jin, Kristen Menou |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 70758 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.15664