Which Cosmos DB vector search metric computes the angle between two vectors and is most commonly used for embedding similarity?
- A Euclidean distance (L2)
- B Dot product
- C Cosine similarity (correct answer)
- D Manhattan distance
Shortcut: Cosine similarity measures the cosine of the angle between two vectors, normalized to ignore magnitude. It is the standard metric for semantic similarity between text embeddings from models like Ada.