01
Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
Patrick Lewis, Ethan Perez, Aleksandra Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, Douwe Kiela
The paper introduces RAG: a model that combines the knowledge stored in a pretrained generator with passages retrieved from an external corpus. This makes generation more specific and factual while allowing the system's knowledge source to be inspected and updated.
What I kept
↗A language model does not need to keep every fact in its parameters—it can retrieve relevant evidence first and generate an answer conditioned on that evidence.
#rag#retrieval#nlp#knowledge-intensive-nlp#open-domain-qa#language-models