Does RAG Really Perform Bad For Long-Context Processing?

Kun Luo, Zheng Liu, Peitian Zhang +3

The efficient processing of long context poses a serious challenge for large language models (LLMs). Recently, retrieval-augmented generation (RAG) has emerged as a promising strat…