1 paper
Zhetong Zhang, Honghao Fu, Miao Xu +2
As Video Large Language Models are increasingly deployed in real-world applications, ensuring their safety alignment has become critical. Counterintuitively, we find that harmful v…