AI engineering glossary · RAG vector search
What is rescoring?
Rescoring (also called Oversampling and rescoring): Re-computing exact or higher-precision similarity for a shortlist found with compressed vectors, then reordering it. It recovers much of the accuracy lost to quantisation while keeping most of the speed benefit.
Rescoring: an example
Binary vectors find 200 candidates, and full-precision vectors rescore them to give the final top 10.
Related terms
More RAG vector search terms
- Retrieval evaluation
- Retrieval-augmented generation
- Retriever
- Scalar quantisation
- Score fusion
- Self-RAG
- Semantic cache
- Semantic chunking
Where you learn it
Rescoring is part of the AI Engineering Program, which teaches it with notes, a hands-on lab and an exam. The first lesson is free.