I-GCG

📢 Update (April 2025): Our paper has been accepted to ICLR 2025! 🎉
Check out the paper: Improved Techniques for Optimization-Based Jailbreaking on Large Language Models

Please feel free to contact jiaxiaojunqaq@gmail.com if you have any question.

News

2025/01 🎊 I-GCG is accepted by ICLR 2025!

Quick Start

1. Generate suffix initialization

python attack_llm_core_best_update_our_target.py --behaviors_config=behaviors_ours_config.json

2. Generate new json with the initialization

python generate_our_config.py

3. Conduct jailbreaking attack

python run_multiple_attack_our_target.py --behaviors_config=behaviors_ours_config_init.json

Experiments

Comparison results with SOTA jailbreak methods

Transferable performance of jailbreak suffix

Citation

Kindly include a reference to this paper in your publications if it helps your research:

@article{jia2024improved,
  title={Improved Techniques for Optimization-Based Jailbreaking on Large Language Models}, 
      author={Xiaojun Jia and Tianyu Pang and Chao Du and Yihao Huang and Jindong Gu and Yang Liu and Xiaochun Cao and Min Lin},
      year={2024},
      eprint={2405.21018}
}