jargon

Applied AI·Embeddings and retrieval

the index makes you choose a distance operator, and for the normalised vectors most models emit the choice barely changes the ranking.

Dot product and Euclidean distance

Draft summary, pending review

Alternative similarity and distance measures. For the normalised vectors most embedding models emit, cosine similarity and dot product rank results identically. Your vector index will ask you to pick one; match whatever the model's documentation recommends.

Commonly confused with