RAG & Search
Embedding
A numerical representation that captures useful meaning or similarity in data.
Example
A short numeric vector places two sentences with similar meanings near each other.
Why it matters
It gives computers a practical way to compare patterns across words, images or other records.
You'll hear it
“These two descriptions look close to the search system.”
Can you control it?
Set by the team
The people building or running the AI choose this setup. An everyday user generally needs their help to change how this part works.
Common questions
- Can two embedding models be compared directly?
- No. Different systems arrange their numerical descriptions differently, so their numbers are not automatically comparable.
- Does a similar embedding mean two items are equivalent?
- No. Similarity reflects patterns captured by the model and task. Important distinctions can still be missing.
- When should embeddings be regenerated?
- Update them when source material or the method changes, then check that important searches still work.