mlsanityData to 14 Oct 2026

SOCIA-EVO: Automated Simulator Construction via Dual-Anchored Bi-Level Optimization

arXiv 2604.17351 ↗First posted 19 Apr 2026cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

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.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 36 of its 63 references that are in the map. First posted on arXiv 19 Apr 2026.

Builds on

36 of 63 references.

References · 36

  1. 2021Evaluating Large Language Models Trained on Codecited by 8,320
  2. 2023Generative Agents: Interactive Simulacra of Human Behaviorcited by 3,568
  3. 2023Lost in the Middle: How Language Models Use Long Contextscited by 3,455
  4. 2023SWE-bench: Can Language Models Resolve Real-World GitHub Issues?cited by 3,373
  5. 2023A Survey on Large Language Model based Autonomous Agentscited by 2,242
  6. 2022Competition-Level Code Generation with AlphaCodecited by 1,462
  7. 2023The Rise and Potential of Large Language Model Based Agents: A Surveycited by 1,439
  8. 2023Is Your Code Generated by ChatGPT Really Correct? Rigorous Evaluation of Large Language Models for Code Generationcited by 1,391
  9. 2023MemGPT: Towards LLMs as Operating Systemscited by 1,389
  10. 2023Large Language Models Cannot Self-Correct Reasoning Yetcited by 1,017
  11. 2023Teaching Large Language Models to Self-Debugcited by 988
  12. 2022Out of One, Many: Using Language Models to Simulate Human Samplescited by 752
24 more
  1. 2024A Survey on the Memory Mechanism of Large Language Model based Agentscited by 634
  2. 2019The frontier of simulation-based inferencecited by 502
  3. 2022InCoder: A Generative Model for Code Infilling and Synthesiscited by 498
  4. 2025Agentic Context Engineering: Evolving Contexts for Self-Improving Language Modelscited by 398
  5. 2023S3: Social-network Simulation System with Large Language Model-Empowered Agentscited by 233
  6. 2024RepairAgent: An Autonomous, LLM-Based Agent for Program Repaircited by 232
  7. 2023Is Self-Repair a Silver Bullet for Code Generation?cited by 197
  8. 2023Who Wrote this Code? Watermarking for Code Generationcited by 141
  9. 2024SWT-Bench: Testing and Validating Real-World Bug-Fixes with Code Agentscited by 133
  10. 2024Large Language Models as Urban Residents: An LLM Agent Framework for Personal Mobility Generationcited by 100
  11. 2024LLM Hallucinations in Practical Code Generation: Phenomena, Mechanism, and Mitigationcited by 93
  12. 2018How Many Random Seeds? Statistical Power Analysis in Deep Reinforcement Learning Experimentscited by 60
  13. 2025Are "Solved Issues" in SWE-bench Really Solved Correctly? An Empirical Studycited by 49
  14. 2024CodeHalu: Investigating Code Hallucinations in LLMs via Execution-based Verificationcited by 34
  15. 2024Serial Position Effects of Large Language Modelscited by 30
  16. 2021DySMHO: Data-Driven Discovery of Governing Equations for Dynamical Systems via Moving Horizon Optimizationcited by 19
  17. 2019Automated Discovery of Business Process Simulation Models from Event Logscited by 17
  18. 2022On learning agent-based models from datacited by 14
  19. 2025AgentSociety Challenge: Designing LLM Agents for User Modeling and Recommendation on Web Platformscited by 12
  20. 2025Hallucination by Code Generation LLMs: Taxonomy, Benchmarks, Mitigation, and Challengescited by 11
  21. 2025G-Sim: Generative Simulations with Large Language Models and Gradient-Free Calibrationcited by 8
  22. 2025MemOrb: A Plug-and-Play Verbal-Reinforcement Memory Layer for E-Commerce Customer Servicecited by 7
  23. 2022BESSIE: A Behavior and Epidemic Simulator for Use With Synthetic Populationscited by 4
  24. 2023Masking Behaviors in Epidemiological Networks with Cognitively-plausible Reinforcement Learningcited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025SOCIA-nabla: Textual Gradient Meets Multi-Agent Orchestration for Automated Simulator Generation
  2. 2026A Closer Look at Agentic BBO: Benchmarking LLM Agents for Black-Box Optimization
  3. 2025ROAD: Reflective Optimization via Automated Debugging for Zero-Shot Agent Alignment
  4. 2026Agentic ESOpt: Fine-Tuning Long-Horizon LLM Agents with Minimal GPU Requirements
  5. 2026BenchTrace: A Benchmark for Testing Reflection Ability and Controlled Evolution in LLM Agents
  6. 2025SOCIA-Nabla: Textual Gradient Meets Multi-Agent Orchestration for Automated Simulator Generation
  7. 2026TTHE: Test-Time Harness Evolution
  8. 2026ARTIS: Agentic Risk-Aware Test-Time Scaling via Iterative Simulation