AI Glossary Term

What does “RAG-Fusion” mean?

Definition

A retrieval method that generates multiple query variants, retrieves results for each, and fuses the rankings.

Browse the full AI Glossary Explore all free AI guides

How RAG-Fusion works

Example

A question about 'laptop battery drain' is rewritten as 'power consumption causes' and 'battery settings tips'. Each version retrieves its own list of documents, and documents that rank well across lists rise to the top.

Why it matters

Several query phrasings cover more of the ways a topic is written about, improving recall. It adds model calls and delay.

Common questions

What is reciprocal rank fusion?

A simple method that gives each document a score based on its rank in every list and adds the scores up.

RAG-Fusion vs multi-query retrieval?

RAG-Fusion adds rank fusion on top of the multi-query idea.

Does it always help?

Not always. Generated variants can drift from the original intent.

RAG (Retrieval-Augmented Generation)A method that retrieves external knowledge and feeds it into generation at inference time.Multi-Query RetrievalA retrieval technique that rewrites the user query into several variants to improve recall.RetrievalFinding relevant documents or records from a knowledge source for a query.Retrieval PrecisionThe proportion of retrieved items that are relevant to the user's query.RoPE (Rotary Positional Embeddings)A positional encoding method that rotates query and key vectors to encode relative token positions.Reinforcement Learning from Human Feedback (RLHF)A training method that uses human preference signals to shape model behavior.

Learn more in our free guides

RAG Stochastic Weight Averaging NaturalSpeech and Latent Diffusion TTS AI in Food and Beverage

Test yourself

RAG Quiz Stochastic Weight Averaging Quiz NaturalSpeech and Latent Diffusion TTS Quiz