- All Courses
- Retrieval-Augmented Generation from Scratch
- Retrieval Quality
- Reranking retrieved candidates
Reranking retrieved candidates
Retrieve broadly, then rerank precisely. A reranker reads query and passage together, which is why it beats vector distance on ordering.
- 5m
- Advanced

Video by Techventures · YouTube
Check your understanding
3 questions. Unlocks when you finish the video.
Overview
What this lesson covers
- Retrieve a wide candidate set, then narrow it
- Score query and passage jointly with a reranker
- Balance the latency cost against the accuracy gain
This lesson sits in Retrieval Quality, part of Retrieval-Augmented Generation from Scratch. It assumes what came before it and leads directly into the next lesson in the module.
In this lesson you will:
- Retrieve a wide candidate set, then narrow it
- Score query and passage jointly with a reranker
- Balance the latency cost against the accuracy gain
Resources
Your notes will live here
Note taking is not available yet. Nothing you type would be saved, so the tab stays read-only for now.