SOCIA-EVO: Automated Simulator Construction via Dual-Anchored Bi-Level Optimization
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910146011070464 |
|---|---|
| author | Hua, Yuncheng Weatherhead, Sion Jafari, Mehdi Xue, Hao Salim, Flora D. |
| author_facet | Hua, Yuncheng Weatherhead, Sion Jafari, Mehdi Xue, Hao Salim, Flora D. |
| contents | Automated simulator construction requires distributional fidelity, distinguishing it from generic code generation. We identify two failure modes in long-horizon LLM agents: contextual drift and optimization instability arising from conflating structural and parametric errors. We propose SOCIA-EVO, a dual-anchored evolutionary framework. SOCIA-EVO introduces: (1) a static blueprint to enforce empirical constraints; (2) a bi-level optimization to decouple structural refinement from parameter calibration; and (3) a self-curating Strategy Playbook that manages remedial hypotheses via Bayesian-weighted retrieval. By falsifying ineffective strategies through execution feedback, SOCIA-EVO achieves robust convergence, generating simulators that are statistically consistent with observational data. The code and data of SOCIA-EVO are available here: https://github.com/cruiseresearchgroup/SOCIA/tree/evo. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_17351 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | SOCIA-EVO: Automated Simulator Construction via Dual-Anchored Bi-Level Optimization Hua, Yuncheng Weatherhead, Sion Jafari, Mehdi Xue, Hao Salim, Flora D. Artificial Intelligence I.2.7 Automated simulator construction requires distributional fidelity, distinguishing it from generic code generation. We identify two failure modes in long-horizon LLM agents: contextual drift and optimization instability arising from conflating structural and parametric errors. We propose SOCIA-EVO, a dual-anchored evolutionary framework. SOCIA-EVO introduces: (1) a static blueprint to enforce empirical constraints; (2) a bi-level optimization to decouple structural refinement from parameter calibration; and (3) a self-curating Strategy Playbook that manages remedial hypotheses via Bayesian-weighted retrieval. By falsifying ineffective strategies through execution feedback, SOCIA-EVO achieves robust convergence, generating simulators that are statistically consistent with observational data. The code and data of SOCIA-EVO are available here: https://github.com/cruiseresearchgroup/SOCIA/tree/evo. |
| title | SOCIA-EVO: Automated Simulator Construction via Dual-Anchored Bi-Level Optimization |
| topic | Artificial Intelligence I.2.7 |
| url | https://arxiv.org/abs/2604.17351 |