让知识库可训练，通过提炼关键信息提升问答准确率
Training the Knowledge Base through Evidence Distillation and Write-Back Enrichment
- 用标注数据定位有效检索结果，提取相关文档精华
- 在六项测试中平均提升2.14%准确率，跨模型也有效
- 适合所有RAG系统，只需一次预处理即可增强知识库
RAG系统中的知识库通常一次性构建后不再更新，尽管查询所需的事实往往分散在多个文档中且混杂在无关内容里。本文认为知识库应作为可训练组件，提出WriteBack-RAG框架：利用标注样本识别检索成功案例，提取相关文档并提炼为紧凑的知识单元，与原始语料一同索引。该方法仅修改语料库，可作为离线预处理步骤，兼容任意RAG流程。在四种RAG方法、六个基准和两个LLM骨干上，WriteBack-RAG均实现提升，平均增益+2.14%。跨方法迁移实验表明，提炼后的知识能惠及非原生生成的RAG系统，证明改进源于语料本身。
原文摘要 · Abstract (English)
The knowledge base in a retrieval-augmented generation (RAG) system is typically assembled once and never revised, even though the facts a query requires are often fragmented across documents and buried in irrelevant content. We argue that the knowledge base should be treated as a trainable component and propose WriteBack-RAG, a framework that uses labeled examples to identify where retrieval succeeds, isolate the relevant documents, and distill them into compact knowledge units that are indexed alongside the original corpus. Because the method modifies only the corpus, it can be applied once as an offline preprocessing step and combined with any RAG pipeline. Across four RAG methods, six benchmarks, and two LLM backbones, WriteBack-RAG improves every evaluated setting, with gains averaging +2.14%. Cross-method transfer experiments further show that the distilled knowledge benefits RAG pipelines other than the one used to produce it, confirming that the improvement resides in the corpus itself.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。