---
title: "AutoSR：通过搜索研究状态实现自动符号回归"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2608.16876v1"
permalink: "https://feed.astralor.com/items/itm_f8870d2930c51e06483a1e41"
published: "2026-08-17T17:55:26.000Z"
score: 76
channel: "ai"
tags: ["arXiv","AutoSR","符号回归","自动科学发现","蒙特卡洛树搜索"]
---

# AutoSR：通过搜索研究状态实现自动符号回归

## AI 点评

> 这篇论文最值得关注的地方在于它把符号回归从单纯找公式提升到模拟科研过程的层面，“研究状态”的保留和复用是很有启发的设计。对做科学发现工具和可解释AI的人来说，这是一个必须跟踪的进展，尤其是它拿下了三个此前无人破解的cp3-bench问题，说明这套思路不是空谈。

## 速览

本文提出AutoSR，一个全自动符号回归系统，通过搜索持久科学调查而非孤立方程来实例化Research-Space Symbolic Regression。传统方法仅优化表达式，丢失研究记录；AutoSR将候选方程与推理、计算证据和独立评审保存在“研究状态”中，并利用渐进式扩展蒙特卡洛树搜索（PW-MCTS）分配计算资源。在9个基准挑战中，AutoSR全部恢复了代数等价关系，包括3个已有系统无法解决的cp3-bench问题。该系统将符号回归从方程级搜索扩展为自动化科学调查，让科学知识与累积证据共同塑造探索方向和最终解释。

## 关键点

- AutoSR在“研究状态”中保存候选方程及其动机、计算证据和独立评审，避免丢失研究记录。
- 采用渐进式扩展蒙特卡洛树搜索（PW-MCTS）跨竞争性调查分配计算资源。
- 在9个基准挑战中全部恢复代数等价关系，包括3个此前无系统解决的cp3-bench问题。
- 将符号回归从方程级搜索扩展为自动化科学调查，强调科学可信度而非单纯数值拟合。

## 全文

We introduce Automatic Symbolic Regression (AutoSR), a fully automated system that instantiates Research-Space Symbolic Regression by searching persistent scientific investigations rather than isolated equations. Finite, noisy data often yield numerically competitive expressions that imply very different behavior outside the observed regime, making numerical fit and syntactic complexity insufficient measures of scientific credibility. Existing approaches largely focus on improving expressions, yet the search typically retains little beyond the resulting formula and score, losing the scientific record, such as motivations and probes, that inform what to try next. AutoSR preserves this record in a \textbf{Research State}, coupling each candidate equation with the reasoning, computational evidence, and independent review developed along its branch. Proposer--reviewer agents develop these states under progressive-widening Monte Carlo tree search (PW-MCTS), which allocates computation across competing investigations, while the accumulated research record is ultimately synthesized into a final report that explains the leading relation and the basis for its selection. Across nine selected challenges from two benchmark suites, AutoSR recovers algebraically equivalent relations in every case, including three cp3-bench problems that no published system recovers and six structurally diverse LSR-Transform problems. Overall, AutoSR extends symbolic regression from equation-level search toward automated scientific investigation, allowing scientific knowledge and accumulated evidence to shape both what is explored and how the resulting equation is justified.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_f8870d2930c51e06483a1e41