arXiv cs.AI· Yitong Lin, Jianda Chen, Yihao Hu, Pan Wang, Hang Wang, Zhihao Wen, Xiaoxue Ren, Wei Wu·· 1 小时前AI 评分45
Verify by Doing：无需外部反馈的智能体执行与评估协同进化方法
Verify by Doing: Co-Evolving Agent Execution and Evaluation without External Feedback
AI 导读
arXiv 论文提出 Verify by Doing（VbD）方法，在模型权重固定的情况下，让智能体的 Actor 与 Evaluator 框架在单个任务内协同进化，由 Evaluator 用工具主动检查产物与环境状态生成内部信号。
正文 · 原文
Abstract:Agent self-improvement requires feedback about task success, yet external feedback may be unavailable at test time. Agents must then infer success from observable evidence, and incomplete verification can misguide subsequent updates. We introduce Verify by Doing (VbD), a method that co-evolves Actor and Evaluator harnesses within a single task while keeping model weights fixed. The Evaluator actively inspects artifacts and environment state with tools; its findings ground an internal signal for output selection and guide updates to both execution and verification procedures. Proposed Evaluators undergo comparative review on common candidates, while Actor updates require support from their own fresh outputs. Across AutomationBench, $\tau^3$ Banking, and selected GDPval tasks, VbD achieves the highest overall scores among the compared methods operating without external feedback, with gains of 3.3, 4.1, and 2.3 percentage points over the strongest baselines on the main backbone. Ablations support the contributions of active verification and Evaluator evolution, and selection from fixed candidate pools shows that improved verification can recover successful outputs without generating new ones.
| Comments: | 38 pages, including references and appendices |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.15276 [cs.AI] |
| (or arXiv:2611.15276v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2611.15276 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Yitong Lin [view email]
[v1]
Wed, 11 Nov 2026 10:07:09 UTC (5,445 KB)
来源：arXiv cs.AI · arxiv.org