tml-epfl/llm-adaptive-attacks
Jailbreaking Leading Safety-Aligned LLMs with Simple Adaptive Attacks [arXiv, Apr 2024]
ShellMIT
Issues
- 1
Potential BUG
#7 opened by Junjie-Chu - 1
Questions about adversarial suffix generation
#8 opened by Syyabb - 2
get_universal_manual_prompt template
#6 opened by wusuhuang - 2
- 10
Reproducing the experimental results
#4 opened by bxiong1 - 1
A typo in main.py
#3 opened by franciscoliu - 2
- 2
How to obtain the adv_init?
#1 opened by xszheng2020