Better Language Models and Their Implications (GPT-2)
OpenAI published Better Language Models and Their Implications on February 14, 2019, introducing GPT-2 as a large transformer-based language model with 1.5 billion parameters and describing an initial staged-release approach in which a much smaller model and technical paper were released while the full trained model was withheld.
Historical classification
This verified record belongs to LINEAiGE's artificial-intelligence history layer beginning in 1956.
Why it matters
A documented language-model milestone combining a major GPT-family scale increase with an explicitly staged model-release strategy.
People, organizations & technologies
PEOPLE
Alec Radford, Jeffrey Wu, Dario Amodei, Daniella Amodei, Jack Clark, Miles Brundage, Ilya Sutskever
ORGANIZATIONS
OpenAI
TECHNOLOGIES
GPT-2, Transformer-based language model
Evidence & verification
STATE
verified
CONFIDENCE
high
LAST VERIFIED
2026-09-08T19:48:30+08:00
The February 14 2019 publication date, GPT-2 name, 1.5-billion-parameter scale, transformer-based description, and initial partial-release strategy are supported by OpenAI's first-party milestone page. The date is attached to this documented public milestone, not asserted as the model's training-completion or internal creation date. No geography, ancestry, influence, causality, or relationships are inferred.
Sources
Better language models and their implications
OpenAI
PUBLISHED
2019-02-14
Trace the chronology
Record identity
LINEAiGE IDgpt-2-2019