Multi-hop retrieval improves question answering by exploring evidence progressively
Beyond One-Shot Expansion: Contrastive Evidence Exploration for Multi-Hop Retrieval
Artificial IntelligenceInformation Retrieval
Summary
Finding the right pieces of information is tricky when answering questions that need several steps, because the clues are linked in complex ways. The authors developed a way to search for information bit by bit, checking what’s already found to guide the next search more smartly. They also use comparisons between similar pieces of information to pick the best clues and cover all parts of the question. Their approach improved results on several challenging multi-step question-answering tests.
What this means in practice
- •For chatbot developers: Enhance chatbot accuracy for complex questions by iteratively retrieving and refining supporting information to cover multiple evidence needs.
- •For enterprise search engineers: Improve retrieval systems by implementing multi-stage exploration of linked documents to better handle queries requiring multiple hops of information.
Authors
JungMin Yun, YoungBin Kim
Abstract
Retrieval-augmented generation (RAG) critically depends on retrieving the evidence necessary for effective reasoning. However, this remains particularly challenging in multi-hop question answering (QA), where supporting passages are often linked through intermediate entities and relations that must be progressively uncovered. Existing retrieval approaches typically rely on a single retrieval intent or one-shot query expansion, limiting their ability to adapt to newly retrieved evidence and potentially introducing noisy or redundant retrieval signals. To address these limitations, we propose a training-free multi-hop retrieval framework that integrates evidence-conditioned exploration, passage-specific contrastive refinement, and coverage-aware final ranking. During offline indexing, the framework constructs passage-specific contrastive facets that characterize each passage relative to its semantically similar neighbors, providing fine-grained signals to distinguish closely related candidates. At inference time, the framework iteratively retrieves evidence, generates probes targeting unresolved information needs, refines candidate relevance using the contrastive facets, and selects a complementary set of passages that collectively cover diverse evidence-seeking intents. Experiments on MuSiQue, HotpotQA, and 2WikiMultihopQA demonstrate consistent improvements in retrieval quality and downstream QA performance over baselines.