A Hitchhiker's Guide to Jailbreaking ChatGPT via Prompt Engineering
Apr 20, 2024·
,,,,,,,·
1 min read
Yi Liu
Gelei Deng
Zhengzi Xu
Yuekang Li
Yaowen Zheng
Ying Zhang
Lida Zhao
Tianwei Zhang
Kailong Wang
Abstract
This work provides a comprehensive guide to jailbreaking ChatGPT through prompt engineering techniques. We systematically categorize jailbreak methods, analyze their effectiveness, and discuss implications for LLM safety and responsible AI deployment.
Type
Publication
4th International Workshop on Software Engineering and AI for Data Quality (SEA4DQ @ ICSE)
This workshop paper provides practical guidance on jailbreaking ChatGPT through prompt engineering, offering insights for both attack and defense research.
