Securing our frontier model training environment
AI Summary
OpenAI explains how we strengthened containment for frontier model training and evaluation, including offline environments and high-compute red-teaming.
Discussion
Sign in to join the discussion.
Loading comments…