- BadToken: Token-level Backdoor Attacks to Multi-modal Large Language Models
In this work, we propose BadToken, the first token-level backdoor attack to MLLMs BadToken introduces two novel backdoor behaviors: Token-substitution and Token-addition, which enable flexible and stealthy attacks by making token-level modifications to the original output for backdoored inputs
- CVPR 2025 Open Access Repository
In this work, we propose BadToken, the first token-level backdoor attack to MLLMs BadToken introduces two novel backdoor behaviors: Token-substitution and Token-addition, which enable flexible and stealthy attacks by making token-level modifications to the original output for backdoored inputs
- BadToken: Token-level Backdoor Attacks to Multi-modal Large Language Models
BadToken: Token-level Backdoor Attacks to Multi-modal Large Language Models （ 2025 03 20 CVPR 25 ） Zenghui Yuan, Jiawen Shi, Pan Zhou1*, Neil Zhenqiang Gong, Lichao Sun Hubei Key Laboratory of Distributed System Security， Hubei Engineering Research Center on Big Data Security，
- dblp: BadToken: Token-level Backdoor Attacks to Multi-modal Large . . .
Bibliographic details on BadToken: Token-level Backdoor Attacks to Multi-modal Large Language Models
- BadToken: Token-level Backdoor Attacks to Multi-modal Large Language Models
Multi-modal large language models (MLLMs) extend large language models (LLMs) to process multi-modal information, enabling them to generate responses to image-t
- Backdoor Vulnerabilities in Multi-modal AI
Research reveals a new token-level backdoor attack method called 'BadToken' that can compromise multi-modal large language models (MLLMs) when processing image-text inputs
- XuankunRong Awesome-LVLM-Safety - GitHub
This survey aims to serve as a cornerstone for future research, facilitating the development of models that not only push the boundaries of multimodal intelligence but also adhere to the highest standards of security and ethical integrity
- 华中科技大学主页平台管理系统 Pan Zhou--英文主页-- Zenghui Yuan, Jiawen Shi, Pan Zhou, etc . . .
Zenghui Yuan, Jiawen Shi, Pan Zhou, etc , “BadToken: Token-level Backdoor Attacks to Multi-modal Large Language Models," Proceedings of the IEEE Conference on Computer Vision
|