Hash-RAG replaces dense vector search with learned binary hash codes for RAG retrieval, reporting a 90% query-time reduction while keeping recall and generation accuracy competitive.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.IR 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
HASH-RAG: Bridging Deep Hashing with Retriever for Efficient, Fine Retrieval and Augmented Generation
Hash-RAG replaces dense vector search with learned binary hash codes for RAG retrieval, reporting a 90% query-time reduction while keeping recall and generation accuracy competitive.