AMIA: Automatic Masking and Joint Intention Analysis Makes LVLMs Robust Jailbreak Defenders.

Yuqi Zhang, Yuchun Miao, Zuchao Li, Liang Ding

Browse the full EMNLP paper archive.