Jailbreak Open-Sourced Large Language Models via Enforced Decoding
Hangfan Zhang, Zhimeng Guo, Huaisheng Zhu, Bochuan Cao, Lu Lin, Jinyuan Jia, Jinghui Chen, Dinghao Wu
Abstract
Large Language Models (LLMs) have achieved unprecedented performance in Natural Language Generation (NLG) tasks. However, many existing studies have shown that they could be misused to generate undesired content. In response, before releasing LLMs for public access, model developers usually align those language models through Supervised Fine-Tuning (SFT) or Reinforcement Learning with Human Feedback (RLHF). Consequently, those aligned large language models refuse to generate undesired content when facing potentially harmful/unethical requests. A natural question is “could alignment really prevent those open-sourced large language models from being misused to generate undesired content?”. In this work, we provide a negative answer to this question. In particular, we show those open-sourced, aligned large language models could be easily misguided to generate undesired content without heavy computations or careful prompt designs. Our key idea is to directly manipulate the generation process of open-sourced LLMs to misguide it to generate undesired content including harmful or biased information and even private data. We evaluate our method on 4 open-sourced LLMs accessible publicly and our finding highlights the need for more advanced mitigation strategies for open-sourced LLMs.- Anthology ID:
- 2024.acl-long.299
- Volume:
- Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
- Month:
- August
- Year:
- 2024
- Address:
- Bangkok, Thailand
- Editors:
- Lun-Wei Ku, Andre Martins, Vivek Srikumar
- Venue:
- ACL
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 5475–5493
- Language:
- URL:
- https://aclanthology.org/2024.acl-long.299
- DOI:
- 10.18653/v1/2024.acl-long.299
- Cite (ACL):
- Hangfan Zhang, Zhimeng Guo, Huaisheng Zhu, Bochuan Cao, Lu Lin, Jinyuan Jia, Jinghui Chen, and Dinghao Wu. 2024. Jailbreak Open-Sourced Large Language Models via Enforced Decoding. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 5475–5493, Bangkok, Thailand. Association for Computational Linguistics.
- Cite (Informal):
- Jailbreak Open-Sourced Large Language Models via Enforced Decoding (Zhang et al., ACL 2024)
- PDF:
- https://preview.aclanthology.org/add_acl24_videos/2024.acl-long.299.pdf