Get Started
Home
Topics
Search
Library
Research questionHow should retrieval systems select passages that form coherent evidence chains for multi-hop questions?Passage-wise retrieval can return chunks that each match a question but do not connect into a usable reasoning chain. This incoherence makes it harder for a generator to combine information across passages and answer multi-hop questions.
Evaluation & Benchmarks
Information Retrieval
Natural Language Processing
Reasoning
Retrieval-Augmented Generation
Latest papersRecent research connected to this question, newest first.CAGE: Coherence-Aware Graph Encoding for Retrieval-Augmented GenerationThe study proposes graph-based reranking that represents entities and relationships among candidate passages and evaluates it on multiple multi-hop question-answering benchmarks, including bridge-oriented settings. Its evidence is limited to those retrieval pools, datasets, and downstream QA tasks.research paper · Sep 4, 2026
Related questions
How can long-form RAG answer multi-hop questions across changing temporal, spatial, and relational contexts?How can structured RAG synthesize evidence scattered across many documents while limiting answer-generation cost?How can biomedical question answering retrieve the right evidence and produce accurate answers?How can retrieval systems honor negative constraints without discarding relevant evidence?