arXiv cs.AI· Letian Huang, Xiangyu Wang, Haowei Lin, Di He·· 2 天前AI 评分22
L2G-SR：学习引导 LLM 符号回归
L2G-SR: Learning to Guide LLM-based Symbolic Regression
AI 导读
L2G-SR 是一个用学习到的结构验证器引导 LLM 符号回归搜索的框架，通过树编辑距离（TED）构建结构监督信号，在搜索时用验证器补充数值误差进行父代选择。在 LLM-SRBench 上，该框架在评测的多个 LLM 骨干和领域上取得数值拟合、方程级准确率和 TED 的 SOTA。
正文 · 原文
Abstract:Symbolic regression searches for an equation that explains observed data. Numerical error is commonly used to guide the search, but it alone can favor an accurate surrogate over a candidate with the right structure. We propose L2G-SR, a framework for LLM-based symbolic regression that guides search with a learned structural verifier. We construct structural supervision signal using tree edit distance (TED), accounting for commutative operators and alternative expression forms. A verifier first aligns numerical observations with the formulas, then learns to score a candidate's structural similarity to the target formula. At search time, the target formula is hidden and the verifier complements numerical error in parent selection. An analyst then uses verifier feedback and diagnostic tools to propose revisions, which a generator converts into refined expressions. On LLM-SRBench, L2G-SR achieves state-of-the-art numerical fit, equation level accuracy and TED across the evaluated LLM backbones and domains. L2G-SR demonstrates the value of learned structural guidance for LLM-based symbolic regression.
| Comments: | 40 pages, including references and appendices |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.20685 [cs.AI] |
| (or arXiv:2610.20685v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.20685 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Letian Huang [view email]
[v1]
Wed, 14 Oct 2026 16:39:39 UTC (867 KB)
来源：arXiv cs.AI · arxiv.org