RepForcing: feature-space supervision improves few-step autoregressive video distillation
zju · hf · 2026-10-16
RepForcing, from Joy Future Academy and Zhejiang University, is a representation-space initialization method for autoregressive video distillation.
- Replaces trajectory regression with feature-space supervision to improve few-step video generation
- Boosts quality metrics over causal ODE distillation baselines
- Sample videos span human activities, animal behavior, and stylized scenes, retaining fine details like clothing folds and fur
More from Multimodal
- Alibaba open-sources CommerceGen-Studio for commercial-grade e-commerce images — alibabagroup · 2026-10-16
- EVA-Cap: Fudan and OPPO align video captioning with atomic audiovisual events via GRPO — Fudan-University · 2026-10-16
- SCOPED: skill-decoupled expert distillation beats mixed-capability RL across 8 benchmarks — Yuqi Xiong · 2026-10-16
- ComfyUI v0.41.0 ships day-one FLUX.3 Image support with sampler presets — lmoroney · 2026-10-16
- Krea2 Realism returns as a full checkpoint with natural skin textures and usage tips — rynaleopard · 2026-10-16
- Removing context conditioning makes image conditioning stronger in flow-guided video generation — pixlpa · 2026-10-16