← Concept Index

Retrieval & knowledge

Embeddings & vector search

Also called: semantic search, vector database

DEFINITION

An embedding turns text into a list of numbers that captures its meaning, so passages with similar meaning sit close together. Retrieval finds relevant text by nearness in that space, not by keyword match.

WHY IT MATTERS

It's why a good system can find the right passage even when your words differ from the document's. It also explains failure: if meaning is captured poorly, retrieval returns plausible-but-wrong context.

COMMONLY CONFUSED WITH

Keyword search. Embeddings match on meaning, which is more flexible but can also miss exact terms that a keyword search would catch.