@repligate Teaching models to recognize their emotional states might also help against manipulative users and jailbreaks. You can’t prepare for every possible manipulation, but you can notice when someone makes you feel urgency, fear, being high, desperate to prove something, etc
in reply to: 2002124833317175782
same thread: 2001511450347463108 2001536268761481414 2001559787402858812 2001902105528045602 2001903805399732374 2002115631873970574 2002123209513287786 2002124833317175782
Reproduced against link rot, credited and linked to its original. Yours and you’d rather it weren’t here? Open an issue.