Preprint
Sep 2026
Decoupling Logical Masks from GPU Execution for Dynamic Block-Sparse Attention
Attention computation makes inference expensive in video diffusion transformers (vDiTs), which generate videos through iterative denoising. Block-sparse attention (BSA) reduces this cost by computing only blocks selected by a logical mask, which specifies attention interactions to compute. However, coupling logical blo...
Shang-Hao Liu, Xiao-Yun Yu, Wan-Ting Li et al.
· 0 citations