AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

Big Bird: Transformers for Longer Sequences

Manzil Zaheer, Guru Guruganesh, Kumar Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, Amr Ahmed

NeurIPS 2020 · Architectures and efficiency › Attention and transformers · approach: Sparse efficient transformers for long context · 267 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
combine random and window attention instead of window only (ablation)sparse attention @512 · SQuADF1: 76.4 -> 85.1Improved
combine random and window attention instead of random only (ablation)sparse attention @512 · MNLIaccuracy: 80.2 -> 80.5Improved
replace full attention with random+window sparse attention (no global) at length 512 (ablation)BERT-base · MNLIaccuracy: 83.4 -> 80.5Hurt
use extended global tokens (etc) instead of internal (itc) (ablation)BigBird base · Natural Questions dev LAF1: 70.8 -> 73.9Improved
use extended global tokens (etc) instead of internal (itc) (ablation)BigBird base · TriviaQA devF1: 79.5 -> 78.7Hurt
longer context via BigBird sparse attention vs RoBERTabase QA model · HotpotQA dev JointF1: 63.5 -> 67.8Improved
BigBird sparse encoder for long inputPegasus large · Arxiv summarizationROUGE-1: 43.85 -> 46.63Improved
BigBird sparse encoder for long inputPegasus large · BigPatent summarizationROUGE-1: 52.25 -> 60.64Improved
increase context from 512 to 4096 with BigBirdDNA MLM · human reference genomeBPC: 1.23 -> 1.12Improved
BigBird vs DeepSeachromatin-profile prediction · DHSAUC: 92.3 -> 92.1No clear effect
BigBird long-context classificationRoBERTa · IMDbaccuracyNo clear effect

Later papers that compared with this method

16 later accepted papers report BigBird as a baseline in their results tables; 12 of them report a better number on most of the results they share with it.

8 more not listed.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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