Defense methods against multi-language and multi-intent LLM attacks

Sunjia Fan; Yichao Yang; Weiqi Huang; Ke Ma; Yucen Liu; Yuxin Zheng

doi:10.1117/12.3051624

18 November 2024 Defense methods against multi-language and multi-intent LLM attacks

Sunjia Fan, Yichao Yang, Weiqi Huang, Ke Ma, Yucen Liu, Yuxin Zheng

Author Affiliations +

Proceedings Volume 13403, International Conference on Algorithms, High Performance Computing, and Artificial Intelligence (AHPCAI 2024) ; 134031L (2024) https://doi.org/10.1117/12.3051624
Event: International Conference on Algorithms, High Performance Computing, and Artificial Intelligence, 2024, Zhengzhou, China

Abstract

This paper proposes a defense framework against jailbreak attacks that exploit multi-language and multi-intent inputs. Research indicates these attacks are effective primarily due to two reasons: (1) LLMs may incorrectly capture key points and semantics in low-resource language inputs, generating malicious content; (2) multiple requests within a single input can cause attention flickering, leading to inadequate capture of implicit requests and incorrect responses. The proposed defense framework requires no additional training and works by mapping multi-language inputs to high-resource languages, guiding the model to think multiple times, decompose intents, and reflect. Experimental results show significant effectiveness of this framework in defending against attacks.

(2024) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Sunjia Fan, Yichao Yang, Weiqi Huang, Ke Ma, Yucen Liu, and Yuxin Zheng "Defense methods against multi-language and multi-intent LLM attacks", Proc. SPIE 13403, International Conference on Algorithms, High Performance Computing, and Artificial Intelligence (AHPCAI 2024) , 134031L (18 November 2024); https://doi.org/10.1117/12.3051624

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

;

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE