RAG & Search
Re-ranking
A second-stage process that reorders retrieved results to put the most relevant items first.
Example
After retrieval, a second model moves the most useful passage to the top.
Why it matters
It puts more attention on choosing the best evidence from an initial shortlist.
You'll hear it
“The useful passage was found, but it needed to be moved higher.”
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 it recover a document missing from the shortlist?
- Not by reordering alone. The document first needs to be included among the candidates.
- Can it add waiting time?
- Yes. Another comparison stage needs extra processing.
- Does first place prove the passage is correct?
- No. Ranking estimates usefulness; it does not verify every factual claim.