mlsanityData to 14 Oct 2026

Hi-Q: Hierarchical Evidence-guided Query Refinement for Multi-Hop Question Answering

arXiv 2608.30468 ↗First posted 31 Aug 2026cs.CL cs.IRcited by 0

Not on the map. Why?

Abstract (arXiv)

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/.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 33 of its 36 references that are in the map. First posted on arXiv 31 Aug 2026.

Builds on

33 of 36 references.

References · 33

  1. 2024The Llama 3 Herd of Modelscited by 14,715
  2. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,297
  3. 2025Qwen3 Technical Reportcited by 8,820
  4. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,695
  5. 2020Dense Passage Retrieval for Open-Domain Question Answeringcited by 3,824
  6. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,754
  7. 2023Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflectioncited by 1,839
  8. 2018Know What You Don't Know: Unanswerable Questions for SQuADcited by 1,554
  9. 2024From Local to Global: A Graph RAG Approach to Query-Focused Summarizationcited by 1,453
  10. 2022Least-to-Most Prompting Enables Complex Reasoning in Large Language Modelscited by 1,407
  11. 2020Constructing A Multi-hop QA Dataset for Comprehensive Evaluation of Reasoning Stepscited by 1,282
  12. 2021Unsupervised Dense Information Retrieval with Contrastive Learningcited by 1,216
21 more
  1. 2022Measuring and Narrowing the Compositionality Gap in Language Modelscited by 1,075
  2. 2022Interleaving Retrieval with Chain-of-Thought Reasoning for Knowledge-Intensive Multi-Step Questionscited by 923
  3. 2021MuSiQue: Multihop Questions via Single-hop Question Compositioncited by 916
  4. 2019ELI5: Long Form Question Answeringcited by 575
  5. 2022Decomposed Prompting: A Modular Approach for Solving Complex Taskscited by 558
  6. 2024RAPTOR: Recursive Abstractive Processing for Tree-Organized Retrievalcited by 532
  7. 2024NV-Embed: Improved Techniques for Training LLMs as Generalist Embedding Modelscited by 472
  8. 2024Adaptive-RAG: Learning to Adapt Retrieval-Augmented Large Language Models through Question Complexitycited by 451
  9. 2024HippoRAG: Neurobiologically Inspired Long-Term Memory for Large Language Modelscited by 281
  10. 2023FinanceBench: A New Benchmark for Financial Question Answeringcited by 270
  11. 2020HybridQA: A Dataset of Multi-Hop Question Answering over Tabular and Textual Datacited by 264
  12. 2025From RAG to Memory: Non-Parametric Continual Learning for Large Language Modelscited by 231
  13. 2021MultiModalQA: Complex Question Answering over Text, Tables and Imagescited by 179
  14. 2020Open Question Answering over Tables and Textcited by 163
  15. 2020MKQA: A Linguistically Diverse Benchmark for Multilingual Open Domain Question Answeringcited by 153
  16. 2020XOR QA: Cross-lingual Open-Retrieval Question Answeringcited by 121
  17. 2024MAIN-RAG: Multi-Agent Filtering Retrieval-Augmented Generationcited by 44
  18. 2025PropRAG: Guiding Retrieval with Beam Search over Proposition Pathscited by 23
  19. 2026Coding Agents are Effective Long-Context Processorscited by 12
  20. 2025Optimizing Question Semantic Space for Dynamic Retrieval-Augmented Multi-hop Question Answeringcited by 6
  21. 2026Retrieval is Cheap, Show Me the Code: Executable Multi-Hop Reasoning for Retrieval-Augmented Generationcited by 5

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026HyperProve: Answer-Guided Hypergraph Expansion for Multi-Hop Question Answering
  2. 2026PAR2-RAG: Planned Active Retrieval and Reasoning for Multi-Hop Question Answering
  3. 2026MEGRAG: Multi-Granular Evidence Graphs for Answer-Aware Multi-Hop RAG
  4. 2025StepChain GraphRAG: Reasoning Over Knowledge Graphs for Multi-Hop Question Answering
  5. 2026Beyond One-Shot Expansion: Contrastive Evidence Exploration for Multi-Hop Retrieval
  6. 2025PRISM: Agentic Retrieval with LLMs for Multi-Hop Question Answering
  7. 2026HyCE-RAG: Hypergraph Chain-of-Evidence Retrieval-Augmented Generation for Explainable Multi-hop Question Answering
  8. 2024Hierarchical Retrieval-Augmented Generation Model with Rethink for Multi-hop Question Answering