Jailbreaking Large Language Models Against Moderation Guardrails via Cipher Characters

Devs

Jailbreaking Large Language Models Against Moderation Guardrails via Cipher Characters | Read Paper on Bytez