cs.CVNov 6, 2026

Trajectory Forcing: Exploiting Diffusion Trajectories for Autoregressive Long Video Generation

Authors: Bohan Wang, Shuo Chen, Yunzhi Li, Zhaozheng Chen, Junzhe Zhang, Hanwang Zhang

Organizations: Huawei Noah Research Institute

Abstract

Autoregressive (AR) modeling is arguably the only way to extend a fixed-length video diffusion model to longer or even streaming videos. However, existing AR diffusion methods use only the previously predicted chunk as context for the next, discarding the intermediate predictions that constitute the chunk's generation history. In contrast, conventional AR models such as LLMs condition on the entire history of preceding tokens. To this end, we propose Trajectory Forcing (TraF), an AR diffusion method that exploits the denoising trajectories of previous video chunks as context. TraF operates under two training regimes: (1) with ground-truth videos, where pseudo-trajectories are constructed from Gaussian interpolations and re-predicted by the model, and (2) with a teacher model, where real trajectories are obtained from the student's own denoising rollout. The first regime naturally provides a strong initialization for the second. On the 5-second VBench benchmark, TraF achieves a Total score of 85.15, outperforming Self Forcing (+0.84) and Causal Forcing (+1.11), with Quality and Semantic improving simultaneously. On 30-second generation---6×\times the training horizon---TraF achieves the highest overall quality among all compared methods.

Figures & tables

Appendix figures & tables2 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. HiAR: Efficient Autoregressive Long Video Generation via Hierarchical Denoising

    Mar 9, 2026Kai Zou, Dian Zheng, Hongbo Liu +3Autoregressive Video GenerationLong-Horizon Video Generation

  2. Clean Forcing: Drift-Resistant Autoregressive Video Diffusion with a Frozen Base

    Oct 19, 2026Wenqing Wang, Joonghyuk Shin, Jonathan Tremblay +2Video Diffusion ModelsAutoregressive Video Generation

  3. Video-Mirai: Autoregressive Video Diffusion Models Need Foresight

    Jun 2, 2026Yonghao Yu, Lang Huang, Runyi Li +2Temporal Consistency in Video GenerationAutoregressive Diffusion