Reciprocal rank fusion
Also called RRF
Combining several retrievers' rankings by summing 1 / (k + rank) for each document, with k typically 60.
Interview questions about Reciprocal rank fusion
- You rank passages with reciprocal rank fusion of BM25 and embeddings, and keep the top one. The passage that answers the query never reaches the model. How?
Fusion rewards agreement, not correctness: each retriever gives 1 / (k + rank), k = 60 being typical.
- How does reciprocal rank fusion work, and why use ranks rather than scores?
Each retriever gives a document 1 / (k + rank); sum across retrievers. BM25 points and cosine similarities aren't on the same scale, but ranks always are.
- Why combine keyword and embedding search, and where does a reranker fit?
Embeddings catch paraphrases; keyword search such as BM25 catches exact terms like product codes, names and error messages.
- Users search an enterprise catalogue of tables, columns and dashboards by exact name and by meaning. How would you design hybrid retrieval over that metadata?
Queries mix exact identifiers, like a table name or catalogue path, with intent, like churn by region.
- Compare tokenization choices and dense versus sparse retrieval for search and RAG, then design and evaluate a hybrid retriever.
Sparse retrieval like BM25 matches tokens, so its tokenization matters: words with stemming suit English prose, while code, part numbers and other languages…
- Design the retrieval and ranking pipeline behind a RAG assistant, including indexing, access control and the metrics you'd evaluate it with.
Index chunks with metadata and their source's access-control list, kept in sync.