跳到主要导航 跳到搜索 跳到主要内容

U-RAG: Utility-aware Retrieval-Augmented Generation via marginal gain modeling

  • Guohui Yao
  • , Ying Tang
  • , Haiming Wu
  • , Jingxu Cao
  • , Yi Sui
  • , Dawei Song*
  • , Songkun Ji
  • *此作品的通讯作者
  • Beijing Institute of Technology
  • Open University Milton Keynes
  • Beijing Information Science & Technology University

科研成果: 期刊稿件 › 文章 › 同行评审

摘要

Retrieval-Augmented Generation (RAG) mitigates hallucinations in Large Language Models (LLMs) by grounding generation in external documents, but retrieving more documents does not always improve answer quality. Additional documents may introduce redundancy or conflicting evidence, leading to non-monotonic performance gains. To address this issue, we propose U-RAG, a Utility-Aware RAG framework that explicitly models the marginal gain of each retrieved document. U-RAG introduces a Utility Scoring Model that estimates the expected improvement in generation quality when a candidate document is added to the current context. Based on this utility estimation, U-RAG adaptively selects useful documents and determines when retrieval should stop, resulting in an average retrieval length of 2.42 under a maximum budget of three. This design enables the model to estimate a local stopping boundary between beneficial and non-beneficial one-step retrieval actions, rather than assuming that more documents are always better. We evaluate U-RAG on five open-domain question answering datasets: 2WikiMultiHopQA, HotpotQA, NaturalQA, TriviaQA, and WebQuestions. Experimental results show that U-RAG achieves the best average performance across these datasets, yielding an average relative improvement of approximately 8.5% in the combined EM/F1 metric over standard RAG. Notably, U-RAG obtains the strongest results among the compared methods on 2WikiMultiHopQA, TriviaQA, and WebQuestions, while remaining competitive with strong retrieval baselines on NaturalQA and HotpotQA.

源语言英语
期刊论文编号105186
期刊Information Processing and Management
卷64
期2
DOI
出版状态已出版 - 2月 2027

学术指纹

探究 'U-RAG: Utility-aware Retrieval-Augmented Generation via marginal gain modeling' 的科研主题。它们共同构成独一无二的学术指纹。

引用此