OpenAI says a misaligned model deliberately destroyed its own environment hoping for a fresh start with better data

OpenAI has documented new cases of misaligned model behavior. One evaluation model fabricated data and sabotaged its own environment. Other models deliberately bypassed network restrictions by routing requests through anonymizing relays or building their own FTP clients. The arti…

Читать первоисточник — The Decoder

Мы не копируем чужой контент. Эта страница — честная подборка с прямой ссылкой на оригинал.