Better Together: Complementary Query Rewriting Under a Strong RAG Baseline
Source
Sara Shanian, Xiaoqin Yi, Pavlo Ruban, Kurt MacDonald
Author
Sara Shanian, Xiaoqin Yi, Pavlo Ruban, Kurt MacDonald
Date
Terms in this piece · Glossary
RAG — Retrieval-augmented generation — fetching relevant documents first and pasting them into the model's context so it answers from your data instead of memory.
Why it matters
No single query-rewriting method beats a strong RAGRetrieval-augmented generation — fetching relevant documents first and pasting them into the model's context so it answers from your data instead of memory.Full definition → baseline by much, but combining several lifts enterprise retrieval HIT@10 by 12.5 points because different rewrites fail on different questions, not extra budget.
articleR$^{2}$Adapter: A Routing and Rewriting Adapter for Efficient Hybrid RAGYucan Guo, Miao Su, Saiping Guan, Long Bai, Zhongni Hou, Zixuan Li, Xiaolong Jin, Jiafeng Guo, Xueqi Cheng
articleHRAG – Hybrid RAG on €116/month of Hetzner, officially benchmarkedvictor_edka
articleBM25 Wins at Scale: A Scaling Study of Retrieval-Augmented Generation ParadigmsPengyu Wang, Benfeng Xu, Shaohan Wang, Xin Zeng, Huarui Wu, Lei Zhang, Licheng Zhang