AI Glossary Term

What does “Bi-Encoder” mean?

Definition

A model that encodes queries and documents into separate vectors so they can be compared quickly at scale.

Browse the full AI Glossary Explore all free AI guides

How Bi-Encoder works

Example

A company embeds every document once overnight. When a user types a query, only the query is embedded and compared against the stored vectors, so results arrive in milliseconds.

Why it matters

Bi-encoders make first-stage retrieval fast at scale. They give up some precision compared with cross-encoders.

Common questions

Why 'bi'?

Query and document are encoded separately, often by the same model.

How is it paired with a cross-encoder?

A bi-encoder fetches candidates quickly and a cross-encoder reranks the short list.

Can I precompute everything?

Yes for documents, which is a main advantage.

EncoderThe component of a model that transforms input into latent representations.Feature ExtractionConverting raw data into informative features that a model can use.Embedding ModelA model specialized for converting data into vectors used for semantic search, clustering, and retrieval.BPE (Byte Pair Encoding)A subword tokenization algorithm that merges the most frequent character pairs into reusable tokens.AGI (Artificial General Intelligence)A hypothetical AI system that can perform most intellectual tasks at a human level across many domains.AI AgentA software system that can observe, reason, and take actions to achieve a goal, often using tools and memory.

Learn more in our free guides

Autoencoders Sparse Autoencoders for Interpretability Cross-Encoders vs Bi-Encoders Sparse Autoencoders for Feature Extraction

Test yourself

Autoencoders Quiz Sparse Autoencoders for Interpretability Quiz Cross-Encoders vs Bi-Encoders Quiz