AI engineering glossary · NLP embeddings
What is cross-encoder?
Cross-encoder: A model that reads two texts together as one input and outputs a single relevance or similarity score. It captures fine interactions between them and is usually more accurate than a bi-encoder, but must be rerun for every pair, so it is slow at large scale.
Cross-encoder: an example
A cross-encoder rescored the top 20 search results for a query to pick the best 3.
Related terms
More NLP embeddings terms
- Cross-lingual transfer
- Dependency parsing
- Distributional hypothesis
- Document-term matrix
- Dot product similarity
- ELMo
- Entity linking
- Entity-level F1
Where you learn it
Cross-encoder is part of the AI Engineering Program, which teaches it with notes, a hands-on lab and an exam. The first lesson is free.