Hybrid Search

1. What is the main advantage of hybrid search compared to pure vector search?

It only uses keyword matching for retrieval. It combines semantic vector similarity with keyword matching. It relies solely on cosine distances for ranking. It simplifies data storage requirements.

2. Which method is primarily used in hybrid search to handle exact identifiers such as error codes or names?

Full-text search Cosine distance Reciprocal Rank Fusion Term-frequency score

3. What is a key reason that pure vector search may fail in certain contexts?

It works well with all types of content. It can accurately score every identifier. It does not prioritize literal tokens or exact matches. It is faster than keyword search.

4. What strategy does LyraLearn use to boost the retrieval of relevant results in hybrid search?

It focuses only on semantic ranking. It ranks chunks by cosine distance then boosts relevant keyword matches. It only retrieves the top results from vector search. It ignores keyword matching entirely.

Continue: When Retrieval Lies — Distance, Similarity, and How to Debug It → Review this lesson Retake quiz