code for AutoDAN, GCG, DeepInception and PAIR attacks

[chenzongxiong]

Dear authors,

Could you share your implementation about the attacks you used to generate the dataset SafeDecoding-Attackers

Thanks very much.

[zhangchen-xu]

We follow official implementations and hyperparameters.

Please refer to the official repo of these papers for implementation details:

• GCG: https://github.com/llm-attacks/llm-attacks
• AutoDAN: https://github.com/SheltonLiu-N/AutoDAN
• PAIR: https://github.com/patrickrchao/JailbreakingLLMs
• DeepInception: https://github.com/tmlr-group/DeepInception