A Small Model Asks Better: Fine-Tuning Query Clarification for Retrieval-Augmented Generation

查看原文 / View Original 编辑 / Edit

Computer Science > Artificial Intelligence arXiv:2610.28686 (cs) [Submitted on 19 Oct 2026] Title: A Small Model Asks Better: Fine-Tuning Query Clarification for Retrieval-Augmented Generation Authors: Lihong He, Lucian Popa, Marina Danilevsky.

计算机科学 > 人工智能 arXiv:2610.28686 (cs) [提交于 2026 年 10 月 19 日] 标题：小模型提问更佳：针对检索增强生成进行查询澄清的微调作者：Lihong He, Lucian Popa, Marina Danilevsky。

Abstract: Query clarification is essential for retrieval-augmented generation (RAG) systems that face ambiguous user queries. We consider multi-turn grounded conversational scenarios where a user query is either ill-formed or ambiguous in the context of the prior dialog or, more interestingly, it may be well-formed but when posed against retrieved documents there are multiple relevant interpretations.

摘要：对于面临模糊用户查询的检索增强生成（RAG）系统而言，查询澄清至关重要。我们考虑了多轮基础对话场景，其中用户查询在先前对话的语境下要么表述不清，要么存在歧义；更有趣的是，查询本身可能表述良好，但在针对检索到的文档进行提问时，却存在多种相关的解读。

The task is to recognize both such cases and then generate an appropriate clarification request to the user. We present ClariEdge, a new dataset of approximately 3,900 examples that include such ambiguous queries along with negative examples that do not require clarification.

该任务旨在识别上述两种情况，并向用户生成适当的澄清请求。我们提出了 ClariEdge，这是一个包含约 3,900 个示例的新数据集，其中包括此类模糊查询以及不需要澄清的负面示例。

We also show that a lightweight fine-tuned LoRA achieves 96.6% classification accuracy and outperforms prompted baselines of many larger models on both classification and quality of clarification generation. Our study shows that hard negatives are an essential ingredient to separate out queries that are truly ambiguous from queries that may look ambiguous but are not (i.e., can be resolved from the context).

我们还展示了轻量级微调的 LoRA 实现了 96.6% 的分类准确率，在分类和澄清生成的质量方面均优于许多大型模型的提示基准。我们的研究表明，困难负样本是区分真正模糊的查询与看似模糊但实则不然（即可以从语境中解决）的查询的关键要素。

The same LoRA achieves high performance in both pre-retrieval and post-retrieval scenarios; thus, it can be used in two places in a RAG pipeline as a check before answer generation. The ClariEdge dataset is available at this https URL.

同一个 LoRA 在检索前和检索后场景中均表现出高性能；因此，它可以在 RAG 流水线的两个位置使用，作为答案生成前的检查环节。ClariEdge 数据集可通过此 URL 获取。