DrAttack: Prompt Decomposition and Reconstruction Makes Powerful LLMs Jailbreakers

Xirui Li, Ruochen Wang, Minhao Cheng, Tianyi Zhou, Cho-Jui Hsieh


Abstract
Safety-aligned Large Language Models (LLMs) are still vulnerable to some manual and automated jailbreak attacks, which adversarially trigger LLMs to output harmful content. However, existing jailbreaking methods usually view a harmful prompt as a whole but they are not effective at reducing LLMs’ attention on combinations of words with malice, which well-aligned LLMs can easily reject. This paper discovers that decomposing a malicious prompt into separated sub-prompts can effectively reduce LLMs’ attention on harmful words by presenting them to LLMs in a fragmented form, thereby addressing these limitations and improving attack effectiveness. We introduce an automatic prompt Decomposition and Reconstruction framework for jailbreaking Attack (DrAttack). DrAttack consists of three key components: (a) ‘Decomposition’ of the original prompt into sub-prompts, (b) ‘Reconstruction’ of these sub-prompts implicitly by In-Context Learning with semantically similar but benign reassembling example, and (c) ‘Synonym Search’ of sub-prompts, aiming to find sub-prompts’ synonyms that maintain the original intent while jailbreaking LLMs. An extensive empirical study across multiple open-source and closed-source LLMs demonstrates that, with fewer queries, DrAttack obtains a substantial gain of success rate on powerful LLMs over prior SOTA attackers. Notably, the success rate of 80% on GPT-4 surpassed previous art by 65%. Code and data are made publicly available at https://turningpoint-ai.github.io/DrAttack/.
Anthology ID:
2024.findings-emnlp.813
Volume:
Findings of the Association for Computational Linguistics: EMNLP 2024
Month:
November
Year:
2024
Address:
Miami, Florida, USA
Editors:
Yaser Al-Onaizan, Mohit Bansal, Yun-Nung Chen
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
13891–13913
Language:
URL:
https://preview.aclanthology.org/icon-24-ingestion/2024.findings-emnlp.813/
DOI:
10.18653/v1/2024.findings-emnlp.813
Bibkey:
Cite (ACL):
Xirui Li, Ruochen Wang, Minhao Cheng, Tianyi Zhou, and Cho-Jui Hsieh. 2024. DrAttack: Prompt Decomposition and Reconstruction Makes Powerful LLMs Jailbreakers. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 13891–13913, Miami, Florida, USA. Association for Computational Linguistics.
Cite (Informal):
DrAttack: Prompt Decomposition and Reconstruction Makes Powerful LLMs Jailbreakers (Li et al., Findings 2024)
Copy Citation:
PDF:
https://preview.aclanthology.org/icon-24-ingestion/2024.findings-emnlp.813.pdf
Software:
 2024.findings-emnlp.813.software.zip