AI Paper Atlas

Research map / Large language models / Retrieval and knowledge

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

NeurIPS 2020 · Large language models › Retrieval and knowledge · approach: Retrieval-augmented reasoning · 18 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add retrieval-augmented generation instead of DPR extractive QANatural Questions · Natural QuestionsExact Match: 41.5 -> 44.5Improved
add retrieval-augmented generation instead of DPR extractive QAWebQuestions · WebQuestionsExact Match: 41.1 -> 45.5Improved
use RAG-Sequence rather than RAG-TokenTriviaQA standard test · TriviaQA (standard test)Exact Match: 55.2 -> 56.8Mixed
use RAG-Token rather than RAG-SequenceJeopardy question generation · Jeopardy question generationQ-BLEU-1: 21.4 -> 22.2Improved
add retrieval-augmented generationMS-MARCO generation versus BART · MS-MARCORouge-L: 38.2 -> 40.8Improved
add retrieval-augmented generationJeopardy question generation versus BART · Jeopardy question generationQ-BLEU-1: 19.7 -> 22.2Improved
add retrieval-augmented generation rather than BARTJeopardy question generation · Jeopardy question generationhuman-rated factuality: 7.1% -> 42.7%Improved
replace learned dense retrieval with fixed BM25 retrieval (ablation)RAG · Natural Questions (dev)Exact Match: 44.0 -> 31.8Hurt
replace learned dense retrieval with fixed BM25 retrieval (ablation)RAG · FEVER 3-way (dev)label accuracy: 74.5 -> 75.1Improved
train by fine-tuning the retriever rather than freezing it (ablation)RAG · Natural Questions (dev)Exact Match: 41.2 -> 44.0Improved
train by fine-tuning the retriever rather than freezing it (ablation)RAG · WebQuestions (dev)Exact Match: 37.1 -> 46.5Improved
add a null-document mechanism (ablation)RAG · Open-domain tasksperformanceNo clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy