#Radial Attention

Radial Attention introduces a novel sparse attention mechanism that cuts training costs by 4.4× and inference time by 3.7× in video diffusion models, enabling generation of longer videos without quality loss.