[论文] [论文] Retriever-Free Retrieval: Navigating Documents with Decision Mode...

论文概要

研究领域: ML 作者: Chao Wan, Paul Jünger, Beixi Du, Bodong Liu, Navaneetha Krishnan et al. 发布时间: 2026-10-26 arXiv: 2610.35868

中文摘要

尽管语言模型的能力已大幅提升，检索仍然主要依赖词法匹配或固定嵌入空间中的相似度，这造成了一种能力不对称：检索器找不到的内容，再强的生成器也无法挽回。我们提出 DecisionRAG，用决策模型将预训练语言模型的能力带入检索——在一次前向传播中联合评估候选分支，全程不生成任何 token；我们还证明，在共享上下文中评判候选，使其相关性估计具有可比性。DecisionRAG 像一个人在图书馆中浏览，沿着语料库固有的结构从宽泛主题一路导航到具体段落，在每一层决定哪些分支值得深入。无需任务特定训练，也无需第一阶段检索器，DecisionRAG 在 MultiDoc2Dial 上以 13.8 个 R@1 点、在 Natural Questions 上以 17.4 个 Hit@1 点，超过配备 8B 重排器的稠密检索。在四个基准（含 268 万维基百科段落的语料库）上，每次查询的检索仅需 0.7–1.8 秒。

原文摘要

While language models have become dramatically more capable, retrieval still largely relies on lexical matching or similarity in fixed embedding spaces. This creates a capability asymmetry: what the retriever fails to find, even the strongest generator cannot recover. We propose DecisionRAG, which brings the capabilities of pretrained language models to retrieval using decision models that jointly evaluate candidate branches in a single forward pass, without generating any tokens, and show that judging candidates in a shared context makes their relevance estimates comparable. Like a person browsing a library, DecisionRAG navigates the corpus's inherent structure from broad topics down to individual passages, deciding at each level which branches are worth exploring. Without task-specific training or a first-stage retriever, DecisionRAG outperforms dense retrieval with an 8B reranker by 13.8 R@1 points on MultiDoc2Dial and 17.4 Hit@1 points on Natural Questions. Across four benchmarks, including a corpus of 2.68 million Wikipedia paragraphs, retrieval takes just 0.7-1.8 seconds per query.

---

*自动采集于 2026-10-28*

#论文 #arXiv #ML #小凯