对齐文档与问题：面向问题的文档重写以增强检索增强生成
Align Documents to Questions: Question-Oriented Document Rewriting for Retrieval-Augmented Generation
AI总结：
本文提出QREAM框架，通过风格控制重写使检索文档更符合问题需求，提升检索增强生成的准确性与效率。
摘要：
检索增强生成（RAG）通过整合检索到的文档和/或生成的上下文来提高大语言模型（LLMs）的事实性。然而，当LLMs面对混合上下文时，往往表现出风格偏见，倾向于选择流畅但虚构的生成内容而非事实性强但结构混乱的检索证据。为解决这一问题，我们提出了QREAM，一种风格可控的重写器，通过问题导向的风格对齐检索文档，保留事实性，使LLM读者更有效地利用信息。我们的框架分为两个阶段：（1）QREAM-ICL，利用风格种子引导迭代重写探索；（2）QREAM-FT，一个从去噪ICL输出中蒸馏出的轻量学生模型。QREAM-FT采用双标准拒绝采样，基于答案正确性和事实一致性进行过滤，确保高质量监督。QREAM无缝集成到现有RAG流程中作为即插即用模块。实验表明，QREAM一致提升了高级RAG流程，产生高达8%的相对改进，且延迟开销极小，有效平衡了问题相关性与事实性。
英文摘要：
Retrieval-Augmented Generation (RAG) enhances the factuality of Large Language Models (LLMs) by incorporating retrieved documents and/or generated context. However, LLMs often exhibit a stylistic bias when presented with mixed contexts, favoring fluent but hallucinated generated content over factually grounded yet disorganized retrieved evidence. This phenomenon reveals that the utility of retrieved information is bottlenecked by its presentation. To bridge this gap, we propose QREAM, a style-controlled rewriter that aligns retrieved documents with a question-oriented style while preserving facts, better for LLM readers to utilize. Our framework consists of two stages: (1) QREAM-ICL, which uses stylistic seeds to guide iterative rewriting exploration; and (2) QREAM-FT, a lightweight student model distilled from denoised ICL outputs. QREAM-FT employs dual-criteria rejection sampling, filtering based on answer correctness and factual consistency to ensure high-quality supervision. QREAM seamlessly integrates into existing RAG pipelines as a plug-and-play module. Experiments demonstrate that QREAM consistently enhances advanced RAG pipelines, yielding up to 8% relative improvement with negligible latency overhead, effectively balancing question relevance with factual grounding.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
