AI 中文总结
A-SR是一种自进化智能体框架，通过分层协调实现符号回归，在LLM-SRBench科学领域和真实世界任务上显著提升了符号回归的Acc@0.01和归一化均方误差指标。
AI 中文摘要
符号回归旨在从数据中发现闭式方程，但现有大语言模型（LLM）引导的方法通常依赖统一的提议循环，将异构搜索失败压缩为标量分数和单个提示。我们提出A-SR，这是一种自进化的智能体框架，将控制单元从表达式编辑转移到角色条件证据视图。A-SR通过协调协议之间的路由、在线评估器-奖励角色策略以及状态路由的过程内存来协调公式发现。在搜索过程中，评估器反馈表征可靠性和生产力，更新角色级效用，并将精英基序、失败轨迹和有效性诊断路由到不同智能体。该框架在两个时间尺度上自进化：在单次运行内，它在不更新LLM参数的情况下调整搜索过程；在多次运行之间，记录的轨迹可以被蒸馏到开源LLM中，作为角色条件提议先验。在LLM-SRBench的四个LSR-Synth科学领域上，使用Llama3.1-8B时，A-SR将Acc@0.01从基线的25.79%提升至48.30%，而A-SR-LoRA将对应的Qwen3-4B结果从24.58%提升至38.29%。在四个真实世界科学发现任务上，A-SR在报告的8个指标中有7个取得了最佳的分布内或分布外归一化均方误差。
英文摘要
Symbolic regression aims to discover closed-form equations from data, but existing LLM-guided methods often rely on a unified proposal loop that compresses heterogeneous search failures into a scalar score and a single prompt. We propose A-SR, a self-evolving agentic framework that shifts the control unit from expression edits to role-conditioned evidence views. A-SR coordinates formula discovery through routing among coordination protocols, an online evaluator-reward role policy, and state-routed process memory. During search, evaluator feedback characterizes reliability and productivity, updates role-level utilities, and routes elite motifs, failure traces, and validity diagnostics to different agents. The framework self-evolves at two timescales: within a run, it adapts the search process without updating LLM parameters; across runs, recorded trajectories can be distilled into open-source LLMs as role-conditioned proposal priors. Averaged over the four LSR-Synth scientific domains in LLM-SRBench, A-SR improves Acc@0.01 over baselines from 25.79% to 48.30% with Llama3.1-8B, while A-SR-LoRA improves the corresponding Qwen3-4B result from 24.58% to 38.29%. On four real-world scientific discovery tasks, A-SR obtains the best in-distribution or out-of-distribution normalized mean squared error on 7 of 8 reported metrics.
Comments18 pages, 8 figures, including appendix
