also: vectoralso: embedding vector
What is Embeddings?
Last updated · 1 min read
The trick: map text into a high-dimensional space where distance means similarity. "Dog" lands near "puppy" and far from "spreadsheet" — even though no programmer ever said so, the geometry is learned from how words are used.
Why you care: semantic search, recommendations and the retrieval step of RAG all run on vector search over embeddings. It is the machinery behind "find documents that mean what I asked", not just ones containing the exact words.