AI Glossary Term

What does “Multi-Query Retrieval” mean?

Definition

A retrieval technique that rewrites the user query into several variants to improve recall.

Browse the full AI Glossary Explore all free AI guides

How Multi-Query Retrieval works

Example

A user asks 'why is my app slow?'. The system also searches 'performance bottlenecks' and 'high latency causes', then merges the results so that relevant pages with different wording are found.

Why it matters

People and documents use different words for the same idea. Multiple queries improve the chance of matching.

Common questions

Who generates the variants?

Typically a language model, prompted to rewrite the question in several ways.

What is the downside?

More searches and more passages, which adds cost and can add noise.

How is it different from hybrid search?

Multi-query varies the query wording; hybrid search varies the search method.

Retrieval PrecisionThe proportion of retrieved items that are relevant to the user's query.RAG-FusionA retrieval method that generates multiple query variants, retrieves results for each, and fuses the rankings.Multi-Head AttentionA transformer mechanism that runs several attention operations in parallel to capture different types of relationships.Intent ClassificationPredicting the user's purpose from a text query to route it correctly.RAG (Retrieval-Augmented Generation)A method that retrieves external knowledge and feeds it into generation at inference time.RetrievalFinding relevant documents or records from a knowledge source for a query.

Learn more in our free guides

Retrieval Quality Multilingual LLMs Multimodal Search Hard Parameter Sharing in Multi-Task Networks

Test yourself

Retrieval Quality Quiz Multilingual LLMs Quiz Multimodal Search Quiz