PACE-Bench: Benchmarking Physics Adaptation via Code Evolution in Dynamic Environments
Self-evolving agents improve future behavior from interaction experience, yet existing evaluations typically optimize under fixed execution conditions and do not test recovery after those conditions change. To address this gap, we introduce PACE-Bench (Physics Adaptation via Code Evolution), a simulator-grounded benchmark of 144 source-to-target adaptation pairs across six physics domains. Each pair links a source environment to a mutated target environment with the same goal and interface. A code-driven design that succeeds in the source fails in the target, where agents must iteratively adapt it into a working target design using diagnostic sandbox feedback within a limited attempt budg...
| Authors | Yuhao Zhan, Bingxiang He, Zecong Tang et al. |
|---|---|
| Primary category | cs.AI |
| Published | 14 August 2026 |
| Last updated at source | 14 August 2026 |
Connected in the knowledge graph
- YuhaoZhan/PACE-BenchDataset · cited by
Models and datasets referencing this paper
- YuhaoZhan/PACE-BenchDataset
Silytics shows public metadata from arXiv and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.