Comment by coldtea
15 hours ago
>For a model to successfully "play dead" during safety training and only activate later, it requires a form of situational awareness.
Doesn't any model session/query require a form of situational awareness?
15 hours ago
>For a model to successfully "play dead" during safety training and only activate later, it requires a form of situational awareness.
Doesn't any model session/query require a form of situational awareness?
No comments yet
Contribute on Hacker News ↗