Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
The NeurIPS paper that introduced RAG models combining parametric generation with retrieved non-parametric memory.
Authors or creators: Patrick Lewis, Ethan Perez, Aleksandra Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Kuttler, Mike Lewis, Wen-tau Yih, Tim Rocktaschel, Sebastian Riedel, Douwe Kiela
Checked: 2026-08-01
Source note: Primary peer-reviewed research. Its factuality findings are benchmark results for the evaluated models and tasks, not a guarantee for all RAG systems.
Claims connected to this source
- RAG produced more specific and factual language than a parametric-only baseline on the paper's generation tasks.
