FD-RAG learns semantic-aware adaptive hypergraphs over local corpora, distills them into compact QA memories, answers covered queries via memory matching, invokes LLM reasoning only when needed, and aggregates anonymized memories federatedly, claiming up to 7.8% higher accuracy and 8.4× lower latenc
arXiv preprint arXiv:2505.00443
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.IR 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
FD-RAG: Federated Dual-System Retrieval-Augmented Generation
FD-RAG learns semantic-aware adaptive hypergraphs over local corpora, distills them into compact QA memories, answers covered queries via memory matching, invokes LLM reasoning only when needed, and aggregates anonymized memories federatedly, claiming up to 7.8% higher accuracy and 8.4× lower latenc