Get Started
Home
Topics
Search
Library
Research questionHow can generative retrieval represent both queries and items for direct matching in existing inverted indexes?LLM-assisted retrieval often expands queries while a separate retriever still performs the matching. This leaves generated representations disconnected from the existing index and limits their direct use for candidate selection.
AI
Information Retrieval
Machine Learning
Natural Language Processing
Reinforcement Learning
Latest papersRecent research connected to this question, newest first.It Takes Two to Match: Co-Evolving Generative Retriever with Reinforcement LearningThe source describes a framework that generates compact keyword representations for both queries and items and matches them through an inverted index. Its evidence covers an internal app marketplace dataset and the public WANDS benchmark, with retrieval F1 as the shared objective.research paper · Sep 1, 2026
Related questions
How can cross-modal generative retrieval avoid hallucinating visual details missing from concise text queries?How can retrieval-augmented generation reason causally over dynamic external information without losing context?How can retrieval systems honor negative constraints without discarding relevant evidence?How can search agents retrieve relevant posts from heterogeneous social feeds for complex queries?