← Back to Glossary
RRF Fusion (Reciprocal Rank Fusion)
Reciprocal Rank Fusion
A formula for merging two ranked result lists (like keyword and vector search) into one combined ranking based on item position.
Why it exists
Dense and sparse searches produce scores on completely different scales. RRF merges them purely based on rank order without score normalization.
How it works
Calculates combined score for each doc: RRF_Score = sum(1 / (k + rank_i)), where k is a constant (typically 60).