How we detect frontier model misalignment
OpenAI uses layered monitoring to detect frontier model misalignment and unauthorized agent behavior so responders can intervene before incidents escalate.
Actus AutomatiséesActus TechAI
How we detect frontier model misalignment
1
0
Leave a Comment