Anthropic apologizes for invisible Claude Fable guardrails

https://www.theverge.com/ai-artificial-intelligence/948280/anthropic-claude-fable-invisible-distillation-guardrail

Comments

8 points · 1 comments · view on lemmy.world

1 Comments

QuizzaciousOtter@lemmy.dbzer0.com · 3 pts · 81d

So the model is both too restricted and already jailbroken at the same time? Wow, what a good job they did.