gpt2-1point5b-params-staged-release
IN premise — entries/2026/06/21/wiki-Generative_pre-trained_transformer.md
Created 2026-06-21T09:50:09+00:00
GPT-2 had 1.5 billion parameters, was released February 2019 with a staged release citing misuse risks, and the full model was released November 2019
Summary
GPT-2 was a 1.5-billion-parameter model that OpenAI initially held back from full public release in February 2019 because they judged it could be misused, then opened it up in November 2019. This sets a concrete baseline for how large a model was at the point where developers started taking misuse risk seriously enough to gate release, which matters for calibrating expectations about model capability thresholds over time.
Dependents
These beliefs depend on this one:
- IN gpt2-foreshadowed-capability-risk-dual-scaling — GPT-2's staged release over misuse concerns, combined with measurements showing 1-7% exact duplicate training data in its outputs, illustrated early tensions between scaling language models and managing associated risks such as memorization and potential misuse.