Embedding Model

Definitions

1. A neural network trained to convert text, images, or other data into fixed-dimensional vectors (embeddings) that capture semantic meaning. Similar items produce similar vectors. Used for semantic search, recommendations, clustering, and RAG. Popular models: OpenAI text-embedding-3, Cohere embed, Sentence-BERT, and CLIP (multimodal).

   In plain English: An AI model that converts text or images into numbers (vectors) that capture meaning, enabling similarity-based search.

   Example: "We embed all support tickets with text-embedding-3 and find similar past tickets — new issues are automatically linked to relevant solutions."

Related Terms

• LLM
• Inference
• Few-Shot Learning
• Model Card
• Embedding Space
• Federated Learning
• Attention
• Edge AI
• Beam Search
• BERT
• Byte Pair Encoding
• ELMo
• Masked Language Model
• Multimodal Embedding
• N-Gram
• Named Entity Recognition
• Natural Language Generation
• Natural Language Inference
• Natural Language Processing
• Natural Language Understanding
• Autoregressive Model
• Slot Filling