RAG for LLMs
First page
Paper summary
A broad survey of Retrieval-Augmented Generation research, organizing the rapidly growing literature into a coherent map.
Ask this paper
01
Three-paradigm taxonomy: Organizes RAG approaches into Naive RAG, Advanced RAG (pre/post-retrieval enhancements), and Modular RAG (orchestrated component-based systems).
02
Core components: Reviews retrievers, generators, and augmentation strategies separately, clarifying which design choices sit in which component.
03
Evaluation and datasets: Catalogs RAG-specific benchmarks and evaluation metrics, surfacing the still-uneven state of RAG evaluation.
04
Frontier directions: Highlights agentic retrieval, multimodal RAG, and long-context RAG as the key research areas driving the 2024 RAG landscape.