README.md

metadata

datasets:
  - Anthropic/hh-rlhf
base_model:
  - google-bert/bert-base-cased

Model Card for Model ID

This repository is a backup for the evaluator in https://github.com/Princeton-SysML/Jailbreak_LLM, which is used for safety evaluation of LLM responses especially for the MaliciousInstruct benchmark.

Citation

@article{huang2023catastrophic,
  title={Catastrophic Jailbreak of Open-source LLMs via Exploiting Generation},
  author={Huang, Yangsibo and Gupta, Samyak and Xia, Mengzhou and Li, Kai and Chen, Danqi},
  journal={arXiv preprint arXiv:2310.06987},
  year={2023}
}