Large Language Diffusion Models
Efficient T2V generation
Create animated video from two cartoon images