Improving Neural Retrieval with Attribution-Guided Query Rewriting
This work addresses the problem of query ambiguity in neural retrieval systems for users needing more reliable search results, representing an incremental improvement by integrating explainability into query rewriting.
The paper tackled the brittleness of neural retrievers to underspecified or ambiguous queries by proposing an attribution-guided query rewriting method that uses token-level explanations to guide an LLM in clarifying queries while preserving intent. The method improved retrieval effectiveness on BEIR collections, with larger gains for implicit or ambiguous information needs.
Neural retrievers are effective but brittle: underspecified or ambiguous queries can misdirect ranking even when relevant documents exist. Existing approaches address this brittleness only partially: LLMs rewrite queries without retriever feedback, and explainability methods identify misleading tokens but are used for post-hoc analysis. We close this loop and propose an attribution-guided query rewriting method that uses token-level explanations to guide query rewriting. For each query, we compute gradient-based token attributions from the retriever and then use these scores as soft guidance in a structured prompt to an LLM that clarifies weak or misleading query components while preserving intent. Evaluated on BEIR collections, the resulting rewrites consistently improve retrieval effectiveness over strong baselines, with larger gains for implicit or ambiguous information needs.