Jailbreaking Black Box Large Language Models in Twenty Queries

Abstract

Uses an attacker LLM to automatically generate jailbreak prompts through iterative refinement achieving high success with only black-box access.

Categories

Framework mappings

@article{chao2025jailbreaking,
  title = {{Jailbreaking Black Box Large Language Models in Twenty Queries}},
  author = {Patrick Chao and Alexander Robey and Edgar Dobriban and Hamed Hassani and George J. Pappas and Eric Wong},
  year = {2025},
  journal = {IEEE SaTML 2025},
  eprint = {2310.08419},
  archivePrefix = {arXiv},
  doi = {10.1109/SaTML64287.2025.00010},
  url = {https://arxiv.org/abs/2310.08419}
}