reranking

The step of re-scoring and reordering candidate documents from an initial search by their relevance to the query to improve retrieval accuracy.

3 articlesLast mentioned

Reranking is the stage in a retrieval system in which candidate documents returned by an initial search are re-scored for relevance to the query and reordered. The initial search uses fast but coarse methods such as vector similarity or keyword matching to collect many candidates, and reranking applies a more precise and expensive model, called a reranker, to evaluate a smaller set carefully. Cross-encoder models that take the query and a document together and output a relevance score are widely used.

In RAG systems, reranking serves as the final step for selecting the documents passed to the LLM, preventing irrelevant documents from entering the context and degrading answers. Commercial and open-source rerankers are available, such as Cohere Rerank and the reranking models in NVIDIA NeMo Retriever.

This entry is based on AIPOST articles and widely known facts. If something is wrong, please send us a correction request.

Articles covering this entry

  • Temperature, Attention and Context Rot: Model Mechanics Behind Agent DesignAI in Practice ·

    Use an embedding model, vector search and a reranker to fetch only relevant chunks.

    Read article
  • Intelligence Plane: Why Memory and Ontology, Not the Model, Define AI AgentsAI in Practice ·

    Continual learning: refinement of components such as retrieval rerankers based on governed human feedback.

    Read article
  • NVIDIA Agent Toolkit Bets on Harness Engineering and Knowledge GraphsAI News ·

    …ed documents, and NVIDIA still invests heavily there: better embedding and reranking models, the argument goes, mean an agent needs fewer search turns to land o…

    Read article

Found a mistake or missing an entry?

Wiki entries are based on AIPOST articles and widely known facts. If an entry is wrong, send us a correction request and we will check and fix it. Names from new articles are added to the wiki after the article is published.