<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="li-etal-2024-drattack">
    <titleInfo>
        <title>DrAttack: Prompt Decomposition and Reconstruction Makes Powerful LLMs Jailbreakers</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Xirui</namePart>
        <namePart type="family">Li</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Ruochen</namePart>
        <namePart type="family">Wang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Minhao</namePart>
        <namePart type="family">Cheng</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Tianyi</namePart>
        <namePart type="family">Zhou</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Cho-Jui</namePart>
        <namePart type="family">Hsieh</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2024-11</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Findings of the Association for Computational Linguistics: EMNLP 2024</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Yaser</namePart>
            <namePart type="family">Al-Onaizan</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Mohit</namePart>
            <namePart type="family">Bansal</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Yun-Nung</namePart>
            <namePart type="family">Chen</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Miami, Florida, USA</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
    </relatedItem>
    <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/.</abstract>
    <identifier type="citekey">li-etal-2024-drattack</identifier>
    <identifier type="doi">10.18653/v1/2024.findings-emnlp.813</identifier>
    <location>
        <url>https://aclanthology.org/2024.findings-emnlp.813/</url>
    </location>
    <part>
        <date>2024-11</date>
        <extent unit="page">
            <start>13891</start>
            <end>13913</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T DrAttack: Prompt Decomposition and Reconstruction Makes Powerful LLMs Jailbreakers
%A Li, Xirui
%A Wang, Ruochen
%A Cheng, Minhao
%A Zhou, Tianyi
%A Hsieh, Cho-Jui
%Y Al-Onaizan, Yaser
%Y Bansal, Mohit
%Y Chen, Yun-Nung
%S Findings of the Association for Computational Linguistics: EMNLP 2024
%D 2024
%8 November
%I Association for Computational Linguistics
%C Miami, Florida, USA
%F li-etal-2024-drattack
%X 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/.
%R 10.18653/v1/2024.findings-emnlp.813
%U https://aclanthology.org/2024.findings-emnlp.813/
%U https://doi.org/10.18653/v1/2024.findings-emnlp.813
%P 13891-13913

Markdown (Informal)

[DrAttack: Prompt Decomposition and Reconstruction Makes Powerful LLMs Jailbreakers](https://aclanthology.org/2024.findings-emnlp.813/) (Li et al., Findings 2024)

• DrAttack: Prompt Decomposition and Reconstruction Makes Powerful LLMs Jailbreakers (Li et al., Findings 2024)

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.