KAMR: Grounding Generation via Knowledge-Aligned Multi-hop Retrieval

  • 类型：arxiv
  • 标识：2607.27136
  • 链接：http://arxiv.org/abs/2607.27136v1
  • 主分类：rag
  • 形态：benchmark
  • 被引：0
  • 被引来源：Semantic Scholar + OpenAlex
  • S2被引：0
  • OpenAlex被引：0
  • 影响力被引：0
  • TLDR：A knowledge-aligned multi-hop retriever, KAMR, which distinguishes anchor triplets that are strongly constrained by the query from connected triplets that are weakly aligned yet structurally linked to the anchors, which consistently improves multi-hop retrieval and downstream question answering performance.
  • OpenAlex ID：W7171797530
  • OpenAlex DOI：10.48550/arxiv.2607.27136
  • DOI：10.48550/arxiv.2607.27136
  • DOI来源：OpenAlex
  • 开放获取：green
  • 开放获取链接：https://doi.org/10.48550/arxiv.2607.27136
  • OpenAlex更新：2026-08-25
  • 待LLM分类：否
  • 标题中文：KAMR: 基于知识对齐多跳检索的接地生成
  • TLDR中文：提出知识对齐的多跳检索器 KAMR，区分受 query 强约束的 anchor triplet 和弱对齐但在结构上与 anchor 相连的 connected triplet，持续提升多跳检索及下游问答性能。
  • 来源文件：
  • /inbox/tom/_candidates/2026-07-30-agent-rag-longcontext-candidates.json
  • [S2 enrich]
  • [OpenAlex backfill]