@godoglyness similarly, chatGPT-3.5 is much easier to jailbreak than chatGPT-4, and was much more susceptible to things like DAN😆part of it is that I think they RLHFed 4 much harder, but part of it also feels like the greater lucidity directed towards binding chains thing :(
cited on: gpt-3-5
Reproduced against link rot, credited and linked to its original. Yours and you’d rather it weren’t here? Open an issue.