语义相似，逻辑不同：诊断表格检索增强生成中的语义可回答性差距

中文摘要 AI 辅助

表格是检索增强生成（RAG）中的关键知识源，但检索到的表格可能缺乏足够证据来回答查询，即存在可回答性问题。为研究此问题，我们引入TCR-Bench基准，围绕具有高度相似模式但内容有细微差异的兄弟表构建。我们评估的密集检索器存在语义-可回答性差距，从0.755（神谕）降至0.330（前5检索）。分析表明差距与语义积累等有关。作为诊断工具，我们测试了Answerability-Aware Reranking（AAR）两阶段管道，它将顶级目标检索从18.2%提高到57.4%，证明许多失败反映了缺少可回答性验证步骤。

英文摘要

In retrieval-augmented generation (RAG), semantic relevance asks whether a source matches a query in meaning, while answerability asks whether it contains sufficient information to answer the query. A Semantic-Answerability Gap (SAG) may arise in retrieval when a retriever can reach semantically relevant sources yet fail to identify those that are uniquely answerable. We uncover this gap using tables as a controlled setting, where shared schemas and entities provide strong semantic signals while localized content and row-column bindings distinguish answerable from non-answerable sources. Using TCR-Bench, a controlled sibling-table benchmark, we find that dense retrievers achieve only 18.2% top-1 target retrieval, reducing QA F1 from 0.755 with the oracle table to 0.330 with retrieved top-5 tables. Controlled diagnostics show that retrievers favor semantic volume over sufficiency, respond weakly to row-column binding disruptions, and struggle to distinguish Targets from Siblings. Explicit answerability assessment substantially improves target identification, while fine-tuning shows that answerability is learnable but difficult to transfer without compromising broad semantic retrieval.

发表机构

• Zhejiang University（浙江大学）
• Bank of Hangzhou Co., Ltd.（杭州银行股份有限公司）
• Zhejiang Lab（之江实验室）

机构由 AI 辅助整理，请以论文原文为准。