Liang Chen, Ghazi Shazan Ahmad, Tianjun Yao, Lingqiao Liu, Zhiqiang Shen: One Last Attention for Your Vision-Language Model. ICCV 2025: 1-10