arXiv cs.AI· Yujia Chen, Yuhui Wang, Junzhe Wang, Shihan Dou, Di Liang, Yujiong Shen, Yuhao Zhou, Zhiheng Xi, Ming Zhang, Senjie Jin, Jiayi Chen, Dingwei Zhu, Yajie Yang, Yuan Hua, Chenyao Lu, Tao Gui, Qi Zhang, Xuanjing Huang·· 2 天前AI 评分38
QREF：通过查询精炼与证据过滤对齐搜索与推理，提升 DeepSearch 智能体
QREF: Improving DeepSearch Agents by Aligning Search with Reasoning via Query Refinement and Evidence Filtering
AI 导读
QREF 是一个即插即用的查询精炼器与证据过滤器，通过按当前推理意图重写搜索查询并保留相关证据，提升 DeepSearch 智能体性能。在四个基准上，WebSailor 的 Pass@1 平均相对提升 33.55%，TongyiDeepResearch 提升 11.04%，且无需修改架构或重训练推理策略。经强化学习训练的查询精炼器将缺乏判别性检索锚点的查询批次比例降低 95%。
正文 · 原文
Authors:Yujia Chen, Yuhui Wang, Junzhe Wang, Shihan Dou, Di Liang, Yujiong Shen, Yuhao Zhou, Zhiheng Xi, Ming Zhang, Senjie Jin, Jiayi Chen, Dingwei Zhu, Yajie Yang, Yuan Hua, Chenyao Lu, Tao Gui, Qi Zhang, Xuanjing Huang
Abstract:DeepSearch agents achieve strong performance on complex information-seeking tasks through iterative search and reasoning. However, they often issue intent-misaligned, under-constrained, or redundant search queries, resulting in poor evidence coverage and accumulating retrieval noise. To mitigate these limitations, we propose QREF, a plug-and-play Query Refiner and Evidence Filter. Specifically, the Query Refiner rewrites search queries according to the current reasoning intent while ensuring complementary coverage across each search batch. The Evidence Filter retains evidence relevant to both the current search batch and the reasoning objective, reducing distracting retrieval results. We initialize the Query Refiner with supervised fine-tuning and optimize it through reinforcement learning based on interactions with a real-world web search environment, reducing ineffective queries. Comprehensive experiments demonstrate that our approach improves the performance of DeepSearch agents across four challenging benchmarks, achieving average relative Pass@1 improvements of 33.55% on WebSailor and 11.04% on TongyiDeepResearch without modifying architectures or retraining their reasoning policies. Compared with the original DeepSearch backbone without QREF, the RL-trained refiner reduces the fraction of query batches that lack discriminative retrieval anchors by 95%. Compared with the SFT-trained refiner, it also lowers the filter-based ineffective query ratio.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.21512 [cs.AI] |
| (or arXiv:2610.21512v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.21512 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Yuhui Wang [view email]
[v1]
Thu, 15 Oct 2026 08:47:44 UTC (759 KB)
来源：arXiv cs.AI · arxiv.org