OpenAI has temporarily paused some advanced reinforcement learning training for new models, as its capabilities rapidly increase. Sam Altman stated that this decision was made to ensure safety and alignment standards keep pace with model development.

Kaynağa git