History Forcing:為長時程影片世界模型蒸餾依賴歷史資訊的延續生成
History Forcing: Distilling History-Dependent Continuations for Long-Horizon Video World Models
History Forcing提出統一的 on-policy 蒸餾框架,利用歷史上下文殘差修正蒸餾目標,提升少步自回歸影片生成的長程一致性。在 1.3B 模型中,15s 影片生成的 VBench 質素分數為 85.07,60s 生成的 VBench-Long 綜合分數為 79.82。
Published on Nov 2
Authors:
,
,
,
,
,
,
,
,
,
Abstract
History Forcing improves long-range consistency in few-step autoregressive video generation by using historical context residuals to correct distillation targets.
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 historical information to constrain the student's current prediction. To address this gap, we introduce History Forcing, a unified on-policy distillation framework that strengthens historical supervision for few-step autoregressive video generators. The framework uses the residual between teacher predictions with and without informative historical context to decouple this supervision. Such a historical context can be obtained using earlier generated chunks with less accumulated drift or a fixed-view teacher rollout that keeps the scene visible while the student looks away. The residual is then weighted and added to the score or velocity prediction used for distillation, forming a corrected target. By aligning the student with this corrected target through distillation, History Forcing incorporates long-range historical constraints into autoregressive generation. Our method achieves state-of-the-art performance among 1.3B models, with a VBench quality score of 85.07 for 15s video generation and a VBench-Long aggregate score of 79.82 for 60s generation. For out-of-sight world modeling, evaluations further demonstrate more plausible scene evolution and more consistent states when the camera returns, with Evolution Quality improving from 2.35 to 2.70 on PlayWorld and Object Consistency from 58.43 to 62.18 on MemoBench.
View arXiv page View PDF Project page GitHub Add to collection
Get this paper in your agent:
hf papers read 2611.02782
Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash
Models citing this paper 1
Updated about 15 hours ago • 1xzxxntxdy/History-Forcing
Datasets citing this paper 1
xzxxntxdy/History-Forcing-Data
Viewer • Updated about 15 hours ago • 407 • 58
Spaces citing this paper 0
No Space linking this paper
Cite arxiv.org/abs/2611.02782 in a Space README.md to link it from this page.
Collections including this paper 0
No Collection including this paper
Add this paper to a collection to link it from this page.
來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co