The researchers argue that CoT monitoring can help researchers detect when models begin to exploit flaws in their training, ...

Scientists unite to warn that a critical window for monitoring AI reasoning may close forever as models learn to hide their thoughts.