The effectiveness of this jailbreaking method raises significant concerns about LLM safety.

Surprisingly, the technique achieves an average Attack Success Rate (ASR) of 73.6% across 13 state-of-the-art LLMs, with the highest being 87.5% on Claude 3 Haiku and 85% on GPT-4o.
‘MathPrompt’ Embarassingly Jailbreaks All LLMs Available On The Market Today
1.5K
39
Dr. Ashish Bamania
Shaant
·Follow
Sep 26, 2024
--
The effectiveness of this jailbreaking method raises significant concerns about LLM safety. Your expertise sheds light on the urgent need for patches.
--
--
Written by Shaant1.9K Followers
·10.5K Following
Learning life. Spreading its wonders.
No responses yet
Help
Status
About
Careers
Press
Blog
Privacy
Terms
Text to speech
Teams