arXiv：cs.AI· Yutong Cheng, Wei Cheng, Wenchao Yu, Dawei Zhou, Haifeng Chen·· 6 小时前
TeSearch：用小型测试套件作为自进化智能体的改进信号
Small but Mighty: Test Suites as Improvement Signals for Self-Evolving Agents
AI 导读
针对自进化智能体难以判断哪些自我改写应被保留的问题，研究者提出 TeSearch——将测试反馈作为可执行信号，用质量-多样性进化算法搜索既小又强的测试套件。在 129 个真实模块和三个骨干模型上，TeSearch 生成的套件质量优于包括 LLM 驱动进化搜索在内的对比方法，且是 LLM 类方法中最小的。
正文
Abstract:A self-evolving agent rewrites its own harness and must decide which rewrites to keep, so its improvement signal is the check a rewrite has to pass. Verifying a rewrite by agent rollout is expensive and subjective, while executable checks such as the repository's own tests are too sparse to catch behavior-changing rewrites, rejecting only a quarter of them. We instead take test feedback as the executable signal, and search for test suites that are both small and strong with a quality-diversity evolutionary algorithm, so that the signal is dense and informative. TeSearch treats suite size as a behavior descriptor, so that quality is pursued at every size and a small, strong suite is reachable, and learns a transition kernel during the search that denoises the LLM as an actuator on that descriptor. Across 129 real-world modules and three backbones, TeSearch delivers the highest-quality suites of any method compared, LLM-driven evolutionary search included, and the smallest of any LLM-based one. Used as the improvement signal of a self-evolving agent, it rejects 59.9% of behavior-changing rewrites against 25.3% for the repository's tests, at a 4.0% false-alarm rate and 3.5 s per decision. The code and data are available at this https URL.
| Comments: | In NeurIPS 2026 Workshop on Self-Evolving Diversity-Driven Search for Robust AI Systems (EvoRobust) |
| Subjects: | Artificial Intelligence (cs.AI); Software Engineering (cs.SE) |
| Cite as: | arXiv:2610.15964 [cs.AI] |
| (or arXiv:2610.15964v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.15964 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Yutong Cheng [view email]
[v1]
Sun, 11 Oct 2026 22:36:54 UTC (214 KB)
来源：arXiv：cs.AI · arxiv.org