검색 결과 재정렬
후보를 관련도 순으로 다시 정렬한다
이 페이지의 본문은 영어로 제공됩니다. 제목과 요약은 한국어로 번역되었습니다.
이 능력이 뜻하는 것
Takes a query and a short list of recalled candidates, and returns a reordered ranking with relevance scores. Its division of labour differs from embedding retrieval: retrieval uses cheap vector neighbours to pull dozens from tens of thousands, whereas reranking applies a costlier model to score those dozens carefully, so it only runs over the candidate set.
기술적으로 구현하는 방법
The typical approach is a cross-encoder: query and candidate are concatenated into one sequence and a single model outputs a relevance score; seeing both at once is more accurate than two-tower encoders, but it cannot be pre-computed and must run per pair. To control latency, a common pattern truncates with a small model first and reranks with a larger one — or simply prompts a large model to score.
대표 제품
4관련 기관
대표적 용도
- Improving evidence quality after RAG retrieval
- Final ranking in e-commerce and content search
- Selecting the most relevant passages for a QA system
- Deduplicating and choosing among candidate answers
성능을 평가하는 방법
- nDCG@k
- Discounted gain in the top-k after reranking, rewarding better placement
- MAP
- Mean of average precision over all relevant items
- MRR
- Quality of the position of the first relevant result
경계와 난점
- Cross-encoders score each candidate separately, so latency climbs steeply with candidate count
- Judgements are unstable on out-of-distribution queries, and truncating long documents causes errors
- It can only reorder what was recalled; anything retrieval missed cannot be recovered