LINEAiGE
AI HISTORY

BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

Jacob Devlin, Ming-Wei Chang, Kenton Lee and Kristina Toutanova submitted BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding on October 11, 2018. The paper introduced BERT, a bidirectional Transformer-based language representation model pretrained from unlabeled text and designed for fine-tuning across downstream natural-language-processing tasks.

October 11, 2018AI HISTORYVERIFIED

Historical classification

This verified record belongs to LINEAiGE's artificial-intelligence history layer beginning in 1956.

Why it matters

A documented language-model milestone introducing BERT and its bidirectional pretraining approach for transfer to multiple natural-language-processing tasks.

People, organizations & technologies

PEOPLE
Jacob Devlin, Ming-Wei Chang, Kenton Lee, Kristina Toutanova

ORGANIZATIONS
Google

TECHNOLOGIES
BERT, Transformer, Language representation pretraining

Evidence & verification

STATE
verified

CONFIDENCE
high

LAST VERIFIED
2026-09-08T06:44:53+08:00

Title, authors, October 11 2018 submission date and the paper's description of BERT as a bidirectional pretrained language representation model are supported by the original arXiv record; Google Research independently identifies the same authors and publication work. No later conference-publication date is substituted for the documented 2018 preprint event.

Sources

SOURCE 1 · PRIMARY · CLAIM-EVIDENCE

BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

arXiv

PUBLISHED
2018-10-11

SOURCE 2 · CORROBORATION

BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

Google Research

Trace the chronology

Record identity

LINEAiGE ID
bert-2018