AI engineering glossary · RAG vector search
What is embedding drift?
Embedding drift: A mismatch that appears when embeddings in an index were made with one model or version and queries with another, or when data changes so much that old vectors no longer represent it well. It degrades retrieval, and the fix is re-embedding.
Embedding drift: an example
After switching embedding models for queries only, recall drops sharply since the index still holds old-model vectors.
Related terms
More RAG vector search terms
- Exact nearest neighbour search
- FAISS
- Faithfulness
- Fixed-size chunking
- Flat index
- Full-text search
- Golden dataset
- GraphRAG
Where you learn it
Embedding drift is part of the AI Engineering Program, which teaches it with notes, a hands-on lab and an exam. The first lesson is free.