Hi-Q：多跳問答中基於階層式證據引導的查詢精煉方法

摘要

多跳問答（QA）的核心瓶頸之一在於：問題表述的粒度，往往與語料庫證據可被檢索的粒度不同。既有方法透過在語料庫上施加固定的圖結構、迭代地重新表述查詢，或是在語料庫上執行一個所生成的程式來處理此種不一致，但這些策略並未明確判斷某一查詢單元何時已受證據支持、何時應被細化。我們將此瓶頸表述為「可檢索粒度發現」，並提出 Hi-Q——一個以證據為條件的階層式查詢細化框架。在每個查詢節點上，解析運算子會檢驗所檢索到的證據是否支持目前的查詢單元；已解析的節點即終止，未解析的節點則由一個保留依賴關係的二元運算子進行擴展，並經由語意涵蓋驗證器檢查。因此，Hi-Q 會建立一棵查詢樹，其拓撲結構由語料庫的支持信號決定，而非由固定的分解模板或預先建構的圖形決定。我們在三個多跳問答基準上評估 Hi-Q，主要是在全語料庫檢索設定下進行；在此設定中，相互依賴的證據必須在開放領域的干擾項之間定位，而非在小型的人工標註池中。在此設定下，Hi-Q 在三個基準上的平均達到 52.3 EM 與 64.0 F1，在該平均上領先迭代檢索基準 IRCoT 15.1 EM / 18.2 F1；在 MuSiQue-full 上，相較於基於圖形的 RAG 基準 PropRAG 領先 11.5 EM / 12.0 F1，且無須進行全語料庫的圖建構。在先前研究所使用的受限支持/干擾設定中，Hi-Q 同樣達到最佳準確度：平均 57.9 EM 與 69.3 F1，領先 PropRAG 5.6 EM / 3.9 F1，並領先 IRCoT 13.7 EM / 15.8 F1。專案頁面可於 https://hi-q-project.github.io/ 取得。

English

A central bottleneck in multi-hop Question Answering (QA) is that the granularity at which a question is expressed often differs from the granularity at which corpus evidence is retrievable. Existing methods address this mismatch by imposing fixed graph structures over the corpus, by iteratively reformulating the query, or by executing a generated program over it, but these strategies do not explicitly decide when a query unit is already supported by evidence and when it should be refined. We formulate this bottleneck as retrievable granularity discovery and introduce Hi-Q, an evidence-conditioned framework for hierarchical query refinement. At each query node, a resolution operator tests whether retrieved evidence supports the current query unit; resolved nodes terminate, while unresolved nodes are expanded by a dependency-preserving binary operator and checked by a semantic coverage verifier. Hi-Q therefore grows a query tree whose topology is determined by corpus support signals rather than by a fixed decomposition template or a pre-built graph. We evaluate Hi-Q on three multi-hop QA benchmarks, primarily under full-corpus retrieval, where dependent evidence must be located among open-domain distractors rather than within a small annotated pool. In this setting Hi-Q reaches 52.3 EM and 64.0 F1 averaged over the three benchmarks, ahead of the iterative retrieval baseline IRCoT by 15.1 EM / 18.2 F1 on that same average, and ahead of the graph-based RAG baseline PropRAG by 11.5 EM / 12.0 F1 on MuSiQue-full, without corpus-wide graph construction. In the restricted supporting/distractor setting used by prior work, Hi-Q likewise attains the best accuracy, with 57.9 EM and 69.3 F1 on average, ahead of PropRAG by 5.6 EM / 3.9 F1 and IRCoT by 13.7 EM / 15.8 F1. The project page is available at https://hi-q-project.github.io/.