论文研究
x.com ·
@omarsar0
IBM提出STAIR：用目录结构改进RAG检索
Elvis介绍IBM的STAIR论文，提出用目录作为生成式检索器的寻址方案，保留文档层级结构。在SearchTome上Recall@1达82.6%，优于微调DSI的76.9%，幻觉率低于0.05%。
AI推荐理由 介绍IBM的STAIR RAG论文，提出用目录结构解决分块问题，提供具体性能数据（Recall@1 82.6%）和低幻觉率，对RAG研究有实际价值。
80
浏览 9,475
点赞 159
评论 14
转发 29
正文
来自IBM的精彩RAG论文。
其中有一些关于如何解决常见RAG问题的非常好的想法。
众所周知，检索器会按长度对长文档进行分块，这丢弃了文档已有的层级结构。
因此，他们提出使用目录。
目录有助于精确编码分块所丢弃的全局结构。
STAIR使用该目录作为生成式检索器的寻址方案，因此模型根据语料库提供的结构，从其自身参数中存储和检索信息。
在SearchTome上，其Recall[@1](https://x.com/1)达到82.6%，而微调的可微搜索索引为76.9%，差距具有统计显著性，BM25为59.5%，DPR为68.7%。
幻觉率保持在0.05%以下，这是对生成式检索的长期反对意见，也是将地址空间基于真实层级结构值得额外结构的原因。消融实验还表明，在训练样本极少的情况下，它也能泛化。
论文：[academy.dair.ai/papers/stair-s…](https://academy.dair.ai/papers/stair-structure-aware-information-retriever-a-novel-dataset-and-llm-based-retrie-2609.03874)
查看原文（中文正文为 AI 翻译）
Great RAG paper from IBM.
There are some really good ideas on how to solve common RAG issues.
It's well known that retrievers chunk long documents by length, which discards the hierarchy the document already has.
So they propose using a table of contents.
A table of contents helps to encodes exactly the global structure that chunking throws away.
STAIR uses that table of contents as the addressing scheme for a generative retriever, so the model stores and retrieves information from its own parameters against a structure the corpus supplies.
On SearchTome, it reaches Recall[@1](https://x.com/1) of 82.6 percent against 76.9 percent for a fine-tuned Differentiable Search Index, a statistically significant gap, with BM25 at 59.5 percent and DPR at 68.7 percent.
Hallucination stays below 0.05 percent, which is the standing objection to generative retrieval and the reason grounding the address space in a real hierarchy is worth the extra structure. The ablations also show it generalizes where very few training samples exist.
Paper: [academy.dair.ai/papers/stair-s…](https://academy.dair.ai/papers/stair-structure-aware-information-retriever-a-novel-dataset-and-llm-based-retrie-2609.03874)
