Weng, F.; Xu, Y.; Fu, C.; Wang, W. MMJ-Bench: A Comprehensive Study on Jailbreak Attacks and Defenses for Vision Language Models. AAAI 2025, 39, 27689-27697.