Functional Homotopy: Smoothing Discrete Optimization via Continuous Parameters for LLM Jailbreak Attacks

Zi Wang, Divyam Anshumaan, Ashish Hooda, Yudong Chen, Somesh Jha

Keywords: alignment, optimization

TL;DR: We introduce the functional homotopy method, which utilizes the continuous parameter space to facilitate the smoothing of the discrete optimization problem associated with tokens.

Abstract: Optimization methods are widely employed in deep learning to address and mitigate undesired model responses. While gradient-based techniques have proven effective for image models, their application to language models is hindered by the discrete nature of the input space. This study introduces a novel optimization approach, termed the *functional homotopy* method, which leverages the functional duality between model training and input generation. By constructing a series of easy-to-hard optimization problems, we iteratively solve these using principles derived from established homotopy methods. We apply this approach to jailbreak attack synthesis for large language models (LLMs), achieving a 20%-30% improvement in success rate over existing methods in circumventing established safe open-source models such as Llama-2 and Llama-3.

Primary Area: alignment, fairness, safety, privacy, and societal considerations

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 8240