Glossary
Embeddings
Numeric vector representations of text or other data that capture meaning.
Definition
An embedding is a fixed-size vector of numbers produced by a model from a piece of data — typically text, but also images, audio, or code — where semantically similar inputs produce mathematically similar vectors. Embeddings power similarity search, clustering, and classification.
Context
Embeddings are the connective tissue of most enterprise AI applications. They turn 'find documents like this' into a fast numeric operation. Quality of embeddings — the model that produced them — has a large effect on retrieval quality, so embedding model selection deserves attention.