VisionReward

Headlines · 2

1. 6d
   Glance Forcing cuts video diffusion distillation to one trajectory

   An arXiv paper (arXiv:2610.20220, submitted 14 Oct 2026) proposes Glance Forcing, a lightweight distillation framework that converts bidirectional diffusion video models into causal few-step autoregressive students. The authors say existing pipelines rely on offline Probability Flow ODE trajectories costing tens of thousands of GPU-hours and terabytes of storage, and that reducing supervision to a single sample causes quality collapse traced to differing curvature in the high-noise and low-noise regimes. Glance Forcing instead decouples the denoising trajectory into slow and fast experts with stage-specific objectives while keeping the pretrained teacher frozen; the paper reports that with only one PF-ODE trajectory it raises VBench Total from 84.04 to 84.22, cuts trajectory-collection cost from thousands of GPU-hours to about one A100 GPU-hour, and shows larger gains in dynamic degree, VisionReward and instruction-following, with human raters consistently preferring it over strong baselines.

   Hugging Face · Papers · 🔥 0
2. 6d
   RepForcing: Representation-Space Initialization for Autoregressive Video Distillation

   The paper proposes Representation Forcing, which learns few-step causal initialization for autoregressive video generators through representation-space supervision from a frozen video diffusion transformer instead of intermediate-teacher trajectory regression; the subsequent asymmetric DMD stage is unchanged. The authors report that at four sampling steps it preserves generation quality better than intermediate-teacher-based initialization before DMD, and that after DMD their four-step chunk-wise model improves VBench Total from 84.04 to 84.66 and VisionReward from 6.326 to 10.074 (+59.25%) over the reported causal-ODE-based baseline. The work is posted as arXiv:2610.20111, submitted on 14 Oct 2026 by Siming Fu, Zheming Fu, Lianrui Mu, Haoyang Huang, Chaojie Gu and Ruizhe He.

   Hugging Face · Papers · 🔥 0

Experience and discussion from the community

Share my VisionReward experienceAsk about VisionReward