Embedding
Expanded definition
Embedding models map inputs to fixed-length vectors used by vector search, clustering, and RAG. Similarity is usually cosine or dot product. Production pipelines care about dimensionality, chunking, multilingual coverage, update cadence, and the cost of re-embedding after a model change. Embeddings are not explanations of meaning; they are geometric indexes that still need filters, reranking, and evaluation.
Related terms
Explore adjacent ideas in the knowledge graph.
Embedding FAQ
What is Embedding?
An embedding is a dense vector that represents text, code, or media so similar items land near each other for search and clustering.
How is Embedding used in AI systems?
Embedding models map inputs to fixed-length vectors used by vector search, clustering, and RAG. Similarity is usually cosine or dot product. Production pipelines care about dimensionality, chunking, multilingual coverage, update cadence, and the cost of re-embedding after a model change.
Related
Comparisons, tools, and models that connect to this idea.