1 paper
Xunzhi Xiang, Yabo Chen, Guiyu Zhang +10
Current autoregressive diffusion models excel at video generation but are generally limited to short temporal durations. Our theoretical analysis indicates that the autoregressive…