Jailbreaking Multimodal Large Language Models via Consistent Cross-Modal Backgrounds.

Meng Yang, Peirou Liang, Zhiqian Wu, Yong Liao

Browse the full TrustCom paper archive.