RL-Index: Reinforcement Learning for Retrieval Index Reasoning

  • 类型:arxiv
  • 标识:2606.16316
  • 链接:http://arxiv.org/abs/2606.16316v1
  • 主分类:rag
  • 形态:method
  • 被引:1
  • 被引来源:Semantic Scholar
  • S2被引:1
  • OpenAlex被引:0
  • 影响力被引:0
  • TLDR:RL-Index is proposed, an indexing framework that formulates retrieval index reasoning as a reinforcement learning problem that consistently improves both retrieval and downstream question-answering performance, while significantly reducing online inference latency.
  • OpenAlex ID:W7164886935
  • OpenAlex DOI:10.48550/arxiv.2606.16316
  • DOI:10.48550/arxiv.2606.16316
  • DOI来源:OpenAlex
  • 开放获取:green
  • 开放获取链接:https://doi.org/10.48550/arxiv.2606.16316
  • OpenAlex更新:2026-07-19
  • 待LLM分类:否
  • 标题中文:RL-Index:面向检索索引推理的强化学习
  • TLDR中文:RL-Index 被提出,是一个将检索索引推理建模为强化学习问题的索引框架,能持续提升检索与下游问答性能,同时显著降低在线推理延迟。
  • 来源文件
  • /inbox/tom/_candidates/2026-06-17-rag-retrieval-reranking-candidates.json
  • [S2 enrich]
  • [OpenAlex backfill]