Skip to main content
Cornell University
We gratefully acknowledge support from the Simons Foundation, member institutions, and all contributors. Donate
arxiv logo > cs > arXiv:2407.10106

Help | Advanced Search

arXiv logo
Cornell University Logo

quick links

  • Login
  • Help Pages
  • About

Computer Science > Software Engineering

arXiv:2407.10106 (cs)
[Submitted on 14 Jul 2024 (v1), last revised 8 Sep 2024 (this version, v4)]

Title:DistillSeq: A Framework for Safety Alignment Testing in Large Language Models using Knowledge Distillation

Authors:Mingke Yang, Yuqi Chen, Yi Liu, Ling Shi
View a PDF of the paper titled DistillSeq: A Framework for Safety Alignment Testing in Large Language Models using Knowledge Distillation, by Mingke Yang and 2 other authors
View PDF HTML (experimental)
Abstract:Large Language Models (LLMs) have showcased their remarkable capabilities in diverse domains, encompassing natural language understanding, translation, and even code generation. The potential for LLMs to generate harmful content is a significant concern. This risk necessitates rigorous testing and comprehensive evaluation of LLMs to ensure safe and responsible use. However, extensive testing of LLMs requires substantial computational resources, making it an expensive endeavor. Therefore, exploring cost-saving strategies during the testing phase is crucial to balance the need for thorough evaluation with the constraints of resource availability. To address this, our approach begins by transferring the moderation knowledge from an LLM to a small model. Subsequently, we deploy two distinct strategies for generating malicious queries: one based on a syntax tree approach, and the other leveraging an LLM-based method. Finally, our approach incorporates a sequential filter-test process designed to identify test cases that are prone to eliciting toxic responses. Our research evaluated the efficacy of DistillSeq across four LLMs: GPT-3.5, GPT-4.0, Vicuna-13B, and Llama-13B. In the absence of DistillSeq, the observed attack success rates on these LLMs stood at 31.5% for GPT-3.5, 21.4% for GPT-4.0, 28.3% for Vicuna-13B, and 30.9% for Llama-13B. However, upon the application of DistillSeq, these success rates notably increased to 58.5%, 50.7%, 52.5%, and 54.4%, respectively. This translated to an average escalation in attack success rate by a factor of 93.0% when compared to scenarios without the use of DistillSeq. Such findings highlight the significant enhancement DistillSeq offers in terms of reducing the time and resource investment required for effectively testing LLMs.
Subjects: Software Engineering (cs.SE)
Cite as: arXiv:2407.10106 [cs.SE]
  (or arXiv:2407.10106v4 [cs.SE] for this version)
  https://doi.org/10.48550/arXiv.2407.10106
arXiv-issued DOI via DataCite
Related DOI: https://doi.org/10.1145/3650212.3680304
DOI(s) linking to related resources

Submission history

From: Mingke Yang [view email]
[v1] Sun, 14 Jul 2024 07:21:54 UTC (1,214 KB)
[v2] Wed, 17 Jul 2024 10:27:17 UTC (1,214 KB)
[v3] Sat, 20 Jul 2024 14:50:34 UTC (1,289 KB)
[v4] Sun, 8 Sep 2024 07:09:33 UTC (1,263 KB)
Full-text links:

Access Paper:

    View a PDF of the paper titled DistillSeq: A Framework for Safety Alignment Testing in Large Language Models using Knowledge Distillation, by Mingke Yang and 2 other authors
  • View PDF
  • HTML (experimental)
  • TeX Source
  • Other Formats
license icon view license
Current browse context:
cs.SE
< prev   |   next >
new | recent | 2024-07
Change to browse by:
cs

References & Citations

  • NASA ADS
  • Google Scholar
  • Semantic Scholar
export BibTeX citation Loading...

BibTeX formatted citation

×
Data provided by:

Bookmark

BibSonomy logo Reddit logo

Bibliographic and Citation Tools

Bibliographic Explorer (What is the Explorer?)
Connected Papers (What is Connected Papers?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)

Code, Data and Media Associated with this Article

alphaXiv (What is alphaXiv?)
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Hugging Face (What is Huggingface?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)

Demos

Replicate (What is Replicate?)
Hugging Face Spaces (What is Spaces?)
TXYZ.AI (What is TXYZ.AI?)

Recommenders and Search Tools

Influence Flower (What are Influence Flowers?)
CORE Recommender (What is CORE?)
  • Author
  • Venue
  • Institution
  • Topic

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)
  • About
  • Help
  • contact arXivClick here to contact arXiv Contact
  • subscribe to arXiv mailingsClick here to subscribe Subscribe
  • Copyright
  • Privacy Policy
  • Web Accessibility Assistance
  • arXiv Operational Status
    Get status notifications via email or slack