Adversarial Training for Multimodal Large Language Models against Jailbreak Attacks

Open in new window