Embedding

An embedding is the numerical representation of words, sentences, or other data as vectors – i.e., arrays of numbers. Similar content receives similar vectors, allowing the AI to mathematically capture meaning. Embeddings are the foundation for search functions, recommendation systems, and RAG: they make it possible to find semantically related content even when different words are used.

Related Terms

• Large Language Model (LLM)
• Retrieval-Augmented Generation (RAG)
• Vector Database

Sources: a16z AI Glossary · CNET AI Terms Glossary

← Back to AI Glossary