RVR: Retrieve-Verify-Retrieve for Comprehensive Question Answering
2026-02-20 • Computation and Language
Computation and LanguageInformation Retrieval
AI summaryⓘ
The authors present a method called retrieve-verify-retrieve (RVR) to find a wide range of answers to questions by searching multiple times. First, a system finds documents related to the question, then another checks which of these documents are good. In the next steps, the question is updated with the verified documents to find even more answers that were missed before. Their method works well even with standard search tools and shows better results than other approaches on several question-answering datasets. This approach helps in getting more complete answers by repeatedly refining the search with verification.
information retrievalmulti-round retrievaldocument verificationquery augmentationanswer coveragerecallQAMPARI datasetout-of-domain datasetsfine-tuning retrieversmulti-answer retrieval
Authors
Deniz Qian, Hung-Ting Chen, Eunsol Choi
Abstract
Comprehensively retrieving diverse documents is crucial to address queries that admit a wide range of valid answers. We introduce retrieve-verify-retrieve (RVR), a multi-round retrieval framework designed to maximize answer coverage. Initially, a retriever takes the original query and returns a candidate document set, followed by a verifier that identifies a high-quality subset. For subsequent rounds, the query is augmented with previously verified documents to uncover answers that are not yet covered in previous rounds. RVR is effective even with off-the-shelf retrievers, and fine-tuning retrievers for our inference procedure brings further gains. Our method outperforms baselines, including agentic search approaches, achieving at least 10% relative and 3% absolute gain in complete recall percentage on a multi-answer retrieval dataset (QAMPARI). We also see consistent gains on two out-of-domain datasets (QUEST and WebQuestionsSP) across different base retrievers. Our work presents a promising iterative approach for comprehensive answer recall leveraging a verifier and adapting retrievers to a new inference scenario.