Материал түпнұсқа тілінде: English.
Securing our frontier model training environment
OpenAI explains how we strengthened containment for frontier model training and evaluation, including offline environments and high-compute red-teaming.
2026 ж. 23 қазан · Дереккөз: OpenAI news
Толық мәтін дереккөзде қолжетімді.