GPT-4 Jailbreaks Itself with Near-Perfect Success Using Self-Explanation

Last updated