ModelRefs / Fusion Retrieval — AI Glossary
Fusion Retrieval — AI Glossary
Combining dense (embedding) and sparse (BM25/keyword) retrieval results using reciprocal rank fusion for improved coverage.
Overview
Fusion retrieval (RAG-Fusion, Hybrid Search) runs the query through both a dense vector index and a sparse BM25/TF-IDF index, then merges result rankings using Reciprocal Rank Fusion (RRF) or learned score fusion. Dense retrieval excels on semantic queries; sparse excels on exact-match and rare-word queries. Hybrid consistently outperforms either alone.
Reference details
| Topic | rag |
|---|---|
| Also known as | hybrid retrieval, hybrid search |
| Last reviewed | 2026-06-24 |
Related terms
Continue your research
Use these connected ModelRefs sections to compare alternatives, inspect implementation paths, and review the evidence and governance boundaries relevant to Fusion Retrieval — AI Glossary.
Frequently asked questions
What is Fusion Retrieval?
Combining dense (embedding) and sparse (BM25/keyword) retrieval results using reciprocal rank fusion for improved coverage.
Is Fusion Retrieval the same as hybrid retrieval?
Yes — hybrid retrieval, hybrid search are common aliases for Fusion Retrieval.
What concepts are related to Fusion Retrieval?
Closely related concepts include dense retrieval, sparse retrieval, mmr.