arXiv：cs.AI· Siru Jiang, Yongzhe Lyu, Shuo Lu, Yubin Wang, Yuxiang Zhang, Yue Liao, Bin Wang, Jian Liang, Tieniu Tan·· 5 小时前AI 评分43
WorldSolver：LLM 智能体能否通过生成求解器模拟物理动态？
WorldSolver: Can LLM Agents Simulate the Physical Dynamics via Solver Generation?
AI 导读
研究者推出 WorldSolver 基准，包含 168 个取自 61 篇经典计算机图形学论文、覆盖 7 个物理领域的仿真任务，要求 LLM 智能体补全求解器代码。评测从执行检查、视觉保真度和物理合理性三个维度展开，GPT-5.6-Sol 与 Claude-Opus-5 表现相对最好，但总分仅 48.7% 和 46.7%。
正文
Abstract:LLM-based agents are increasingly advancing scientific and engineering problem solving, with physics simulation emerging as a challenging yet practical testbed for reproducing complex physical phenomena with application in embodied AI, games and films. As the workhorse of such simulation, a solver computes how the state of a dynamic system evolves over time. Building such solvers requires physical understanding to identify appropriate models, mathematical reasoning to formulate the underlying dynamics, and software engineering to implement them as executable code, yet this capability of LLM agents remains underexplored. To this end, we introduce WorldSolver, a benchmark of 168 simulation tasks derived from physical phenomena in 61 classic computer graphics papers, spanning 7 physical domains. Each task contains a code scaffold that provides a fixed simulation environment for the scene, with the solver implementation left for the agent to complete. Specifically, we evaluate them along three dimensions: Execution Checks for successful execution, Visual Fidelity for reproducing the intended dynamic behavior in the rendered simulation, and Physical Plausibility for physics-grounded verification of the generated dynamics. Experiments on frontier agents reveal that producing executable solvers is difficult itself, and satisfying visual and physical correctness is even harder. GPT-5.6-Sol and Claude-Opus-5 perform comparatively better than the other evaluated agents, yet achieve overall scores of only 48.7% and 46.7%, respectively. WorldSolver is an early step toward agentic solver generation, and we hope it helps drive progress toward agents that can faithfully simulate the dynamic physical world. Code is available at this https URL.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.08720 [cs.AI] |
| (or arXiv:2610.08720v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.08720 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Siru Jiang [view email]
[v1]
Tue, 6 Oct 2026 17:26:26 UTC (5,204 KB)
来源：arXiv：cs.AI · arxiv.org