arXiv cs.AI· Haoyang Zhang, Qumeng Sun, Xiao Zhang, Xiang Liu, Zhidong Fan, Junjie Wang·· 11 小时前AI 评分41
Evolver：将智能体探索转化为可复用进化能力
Evolver: Turning Agent Exploration into Evolving Capabilities
AI 导读
Evolver 是一个经验驱动框架，通过结构化 Gene 资产整合经验获取、检索与反馈引导进化，在不更新基础模型参数的前提下让智能体探索经验持久复用。
正文 · 原文
Abstract:Agent exploration produces valuable procedural knowledge, yet much of this experience remains confined to transient execution traces and is not systematically reused. We introduce Evolver, an experience-driven framework that transforms agent exploration into persistent, reusable capabilities without updating foundation model parameters. Built around structured Gene assets, Evolver integrates experience acquisition, retrieval, and feedback-guided evolution to support experience-conditioned execution across existing agent harnesses. We evaluate Evolver using Claude Code and Codex, two foundation model configurations, and four benchmarks: Terminal-Bench 2.1, SWE-bench Multilingual, SWE-rebench, and GAIA. On a selected diagnostic set of 320 paired executions, Evolver improves task success from 83.13% to 87.50% (+4.38 percentage points) while reducing recorded task-execution token consumption by 24.24%. Furthermore, among 265 pairs successful under both conditions, token consumption decreases by 17.57%, demonstrating efficiency benefits even when task success is preserved. These findings suggest that persistent procedural experience can improve both task effectiveness and execution efficiency, providing a practical foundation for transforming agent exploration into evolving capabilities.
| Comments: | 12 pages, 5 figures |
| Subjects: | Software Engineering (cs.SE); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.07595 [cs.SE] |
| (or arXiv:2611.07595v1 [cs.SE] for this version) | |
| https://doi.org/10.48550/arXiv.2611.07595 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Zhang Xiao [view email]
[v1]
Thu, 5 Nov 2026 07:44:28 UTC (1,504 KB)
来源：arXiv cs.AI · arxiv.org