Abstract
Autoregressive (AR) video diffusion enables low-latency, streamable video generation, but prediction errors often accumulate over long rollouts. Training the generator on its own rollouts exposes it to these imperfect histories. However, existing video-level distribution matching distillation (DMD) scores the whole rollout jointly. Because a chunk is evaluated together with its past and future, its correction can favor matching artifacts in the surrounding context merely to preserve temporal consistency. To provide a clearer visual-quality signal, we introduce Rollout-Marginal Distillation (RMD). RMD retains the generated history for AR prediction but scores each chunk independently against a chunk teacher, ensuring its quality correction is not compromised by an imperfect temporal context. To compensate for the lack of temporal context in independent chunk scoring, RMD subsequently applies video-level DMD to restore temporal coherence. Extensive experiments demonstrate that RMD maintains high visual quality far beyond its training horizon and outperforms video-level DMD baselines. Code and video results are available at https://cjeen.github.io/RMD
Keywords
Subject
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
Gao, C., Hu, Z., Ma, J., Zhang, J., Zhang, W., & Xue, T. (2026). Rollout-Marginal Distillation for Long-Horizon Autoregressive Video Generation. https://omanscience.com/en/articles/rollout-marginal-distillation-for-long-horizon-autoregressive-video-generation
MLA 9
Gao, Chenjian, et al. "Rollout-Marginal Distillation for Long-Horizon Autoregressive Video Generation." https://omanscience.com/en/articles/rollout-marginal-distillation-for-long-horizon-autoregressive-video-generation.
Chicago (author–date)
Gao, Chenjian, Zhihao Hu, Jianqi Ma, Jun Zhang, Weidong Zhang, and Tianfan Xue. 2026. "Rollout-Marginal Distillation for Long-Horizon Autoregressive Video Generation." https://omanscience.com/en/articles/rollout-marginal-distillation-for-long-horizon-autoregressive-video-generation.
Harvard
Gao, C., Hu, Z., Ma, J., Zhang, J., Zhang, W. and Xue, T. (2026) 'Rollout-Marginal Distillation for Long-Horizon Autoregressive Video Generation', Available at: https://omanscience.com/en/articles/rollout-marginal-distillation-for-long-horizon-autoregressive-video-generation.
Vancouver
Gao C, Hu Z, Ma J, Zhang J, Zhang W, Xue T. Rollout-Marginal Distillation for Long-Horizon Autoregressive Video Generation. https://omanscience.com/en/articles/rollout-marginal-distillation-for-long-horizon-autoregressive-video-generation
IEEE
C. Gao, Z. Hu, J. Ma, J. Zhang, W. Zhang, and T. Xue, "Rollout-Marginal Distillation for Long-Horizon Autoregressive Video Generation," https://omanscience.com/en/articles/rollout-marginal-distillation-for-long-horizon-autoregressive-video-generation.