Skip to content
Preprint

RepForcing: Representation-Space Initialization for Autoregressive Video Distillation

Oct 2026 · 0 citations · 38 references
Computer Science

TL;DR

This work proposes Representation Forcing, which learns few-step causal initialization through representation-space supervision rather than intermediate-teacher trajectory regression, and shows that, at four sampling steps, RepForcing better preserves generation quality than intermediate-teacher-based initialization before DMD.

Abstract

Distilling pretrained video diffusion models into few-step autoregressive (AR) generators typically combines student initialization with distribution matching distillation (DMD). An intermediate AR teacher can provide causally compatible targets for ODE initialization. However, causally compatible targets alone do not ensure high-quality generation under a short sampling schedule. We analyze the posterior-mean predictions favored by direct latent regression and empirically find that causal ODE initialization alleviates blur but still leaves fine detail unresolved. To address this limitation, we propose Representation Forcing, which learns few-step causal initialization through representation-space supervision rather than intermediate-teacher trajectory regression. The student is trained with teacher forcing by comparing clean predictions and training targets through normalized features of a frozen video diffusion transformer. This phase reuses the synthetic videos generated by a bidirectional model for AR teacher training and replaces both teacher training and causal ODE distillation, without generating additional causal ODE trajectories. The subsequent asymmetric DMD procedure remains unchanged. Empirical results show that, at four sampling steps, RepForcing better preserves generation quality than intermediate-teacher-based initialization before DMD. After DMD, our four-step chunk-wise model improves VBench Total from 84.04 to 84.66 (+0.62 points) and VisionReward from 6.326 to 10.074 (+59.25%) over the reported causal-ODE-based baseline.

View source

Similar papers

Preprint Oct 2026

Direct Forcing: Single-Stage Training with Representation-Space Supervision for Few-Step Autoregressive Video Generation

Few-step autoregressive video generation enables low-latency streaming, yet prevailing training pipelines separate causal adaptation from distribution matching. Recent representation-based text-to-video methods retain this separation by starting with an ODE initialization stage, even when the target distribution is def...

Hao-Jun Xu, Jie Huang, Chu-Yang Zhao et al. · 0 citations
Sep 2026

ViRDM: Taming Representation Distribution Matching for Few-Step Causal Video Generation

ViRDM, a teacher- and critic-free video post-training recipe that turns three-network distillation into generator-only post-training, reducing GPU memory use and training time while improving video quality, is introduced.

Zi-Chong Meng, Chong-Jian Ge, C. Huang et al. · 2 citations · ⚡2

From Scores to Samples: Elastic Forcing for Autoregressive Video Generation

This framework minimizes maximum mean discrepancy in frozen self-supervised video representation spaces in frozen self-supervised video representation spaces, using a hybrid Nystr\"om--Monte Carlo estimator to balance approximation bias and sampling variance.

Chi Zhang, Yue-Yi Liu, Hao-Yan Shi et al. · 3 citations · ⚡2
Preprint Oct 2026

Learning the Autoregressive: Direct Score-Difference Estimation with Amortized Future Credit for Video Distillation

Few-step distillation for autoregressive (AR) video generation commonly first initializes an AR student and then refines it with distribution matching distillation (DMD). This refinement process faces two challenges. First, although initialization already yields a reasonably capable student, DMD obtains the refinement...

Zi-Xuan Duan, Ya-Bo Chen, Xun-Zhi Xiang et al. · 0 citations
Preprint Nov 2026

History Forcing: Distilling History-Dependent Continuations for Long-Horizon Video World Models

Few-step autoregressive video diffusion enables efficient generation, yet maintaining consistency over long rollouts remains challenging, especially when subjects or scenes reappear after going out of sight. Despite progress in fidelity and rollout robustness, existing distillation methods do not fully exploit richer h...

Hang-Yi Kuang, Cheng-Zhi Wang, Ye-Fei He et al. · 0 citations
Sep 2026

Enhancing Autoregressive Video Generation via Representation Adversarial Distillation

Few-step autoregressive video generation enables efficient streaming synthesis, but errors introduced in early temporal blocks are reused as context and can propagate through subsequent rollouts, leading to detail degradation, structural drift, and unstable motion. Existing distribution matching distillation (DMD) prim...

Fang-Yu Lin, Xingtong Ge, Lu Zhu et al. · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.