arXiv：cs.AI· Pengze Li, Jiaquan Zhang, Encheng Su, Yunbo Long, Zihang Zeng, Xinping Liu, Han Zhang, Junchi Yu, Shixiang Tang, Aoran Wang, Xi Chen·· 4 小时前
耦合式工具-模型协同进化框架：用可验证误差信号同时优化科学方程发现系统的工具库与模型策略
Coupled Harness--Model Evolution for Scientific Equation Discovery
AI 导读
研究者提出耦合进化框架，从同一可验证拟合误差信号出发，同时改进可执行工具库与模型工具使用策略：工具进化阶段智能体通过诊断自身失败来新增工具并修正选择规则，策略进化阶段用冻结工具库收集的成功轨迹对约 3B 激活参数的小开源模型进行微调。
正文
Authors:Pengze Li, Jiaquan Zhang, Encheng Su, Yunbo Long, Zihang Zeng, Xinping Liu, Han Zhang, Junchi Yu, Shixiang Tang, Aoran Wang, Xi Chen
Abstract:Large language models (LLMs) are increasingly used to discover scientific equations from experimental data. Such a discovery system has two parts that must work together: a library of executable computational tools and the model's tool-use policy. The tools supply the solving capability, from diagnosing structure in the data to fitting candidate equations; the policy plans and orchestrates their use. Yet existing systems adapt only one part, keeping either the toolset or the model fixed, which leaves the corresponding failures uncorrectable. We develop a coupled evolution framework that improves both parts from the same verifiable fit-error signal. In the tool-evolution phase, an agent grows an executable tool library from a small set of generic numerical operations by diagnosing its own failures: it authors a new tool when a capability is missing and revises its selection rules when the right tool went unused. In the policy-evolution phase, a small open model (roughly 3B active parameters) is fine-tuned on successful tool-use trajectories collected with the frozen library, on training formulas screened to be disjoint from all evaluation benchmarks. Ablations show the two phases are complementary: with no tool library at all the same model solves almost nothing, the tools account for most of the gain, and trajectory training adds a further gain that is largest on the most difficult problems. The full system solves 79.1/71.3\% of LSR-Synth problems at fit tolerances $10^{-2}$/$10^{-3}$, reaches 63.8\% pooled accuracy on the distributionally different SRSD-Feynman panel, and obtains a normalized CP3-28 aggregate of $-7.27$. These results suggest a broader recipe for scientific agents: let the model and its instruments improve each other under verifiable feedback.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.17740 [cs.AI] |
| (or arXiv:2610.17740v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.17740 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Pengze Li [view email]
[v1]
Tue, 13 Oct 2026 05:17:45 UTC (420 KB)
来源：arXiv：cs.AI · arxiv.org