Research map / Video / Video generation
Causal Forcing: Autoregressive Diffusion Distillation Done Right for High-Quality Real-Time Interactive Video Generation
Hongzhou Zhu, Min Zhao, Guande He, Hang Su, Chongxuan Li, Jun Zhu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use AR teacher (causal ODE initialization) instead of bidirectional teacher for ODE distillation, chunk-wise (ablation) | Causal Forcing (ODE init + DMD) · 100-prompt motion set | Dynamic Degree: 24 -> 68 | Improved |
| use AR teacher (causal ODE initialization), chunk-wise (ablation) | Causal Forcing · 100-prompt motion set | VisionReward: 3.330 -> 6.326 | Improved |
| use AR teacher (causal ODE initialization), frame-wise (ablation) | Causal Forcing · 100-prompt motion set | Dynamic Degree: 2 -> 64 | Improved |
| use AR teacher (causal ODE initialization), frame-wise (ablation) | Causal Forcing · 100-prompt motion set | VisionReward: 1.951 -> 6.204 | Improved |
| train AR teacher with teacher forcing instead of diffusion forcing (ablation) | AR diffusion training · 100-prompt motion set | VisionReward: 1.583 -> 3.343 | Improved |
| train AR teacher with teacher forcing instead of diffusion forcing (ablation) | AR diffusion training · 100-prompt motion set | Dynamic Degree: 60 -> 50 | Hurt |
| use causal CD with AR teacher instead of asymmetric CD (ablation) | Consistency distillation · VBench Total | Total score: 79.07 -> 81.48 | Improved |
| apply standard DMD only for AR student initialization (no causal ODE init) (ablation) | Self Forcing · architectural gap isolation | Dynamic Degree / Instruction Following | Hurt |
| Causal Forcing versus Self Forcing at same latency | Distilled AR video models · 100-prompt motion set | Dynamic Degree: 57 -> 68 | Improved |
Later papers that compared with this method
13 later accepted papers report Causal Forcing as a baseline in their results tables; 13 of them report a better number on most of the results they share with it.
- Attend Locally, Remember Linearly: Linear Attention as Cross-Frame Memory for Autoregressive Video Diffusion (NeurIPS 2026 · better on 4 of 6 shared results)
- AR-CoPO: Align Autoregressive Video Generation with Contrastive Policy Optimization (ECCV 2026 · better on 4 of 5 shared results)
- PhyMotion: Structured 3D Motion Reward for Physics-Grounded Human Video Generation (NeurIPS 2026 · better on 3 of 4 shared results)
- What Remains in Sight? Autoregressive Video Decoding as Representation-Guided Context Rewriting (NeurIPS 2026 · better on 4 of 4 shared results)
- Beyond Distribution Matching: Self-Supervised Representation Forcing for Few-Step Video Generation (NeurIPS 2026 · better on 4 of 4 shared results)
- HiAR: Efficient Autoregressive Long Video Generation via Hierarchical Denoising (ECCV 2026 · better on 4 of 4 shared results)
- Disen-Forcing: Disentangling Semantic Anchoring from Motion for Autoregressive Video Diffusion (NeurIPS 2026 · better on 3 of 3 shared results)
- AdaState: Self-Evolving Anchors for Streaming Video Generation (NeurIPS 2026 · better on 3 of 3 shared results)
5 more not listed.
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- From Slow Bidirectional to Fast Autoregressive Video Diffusion Models (CVPR 2025 · similarity 0.91)
- Self Forcing: Bridging the Train-Test Gap in Autoregressive Video Diffusion (NeurIPS 2025 · similarity 0.91)
- AR-Diffusion: Asynchronous Video Generation with Auto-Regressive Diffusion (CVPR 2025 · similarity 0.90)
- Taming Teacher Forcing for Masked Autoregressive Video Generation (CVPR 2025 · similarity 0.89)
- Video-Mirai: Autoregressive Video Diffusion Models Need Foresight (NeurIPS 2026 · similarity 0.89)
- Ca2-VDM: Efficient Autoregressive Video Diffusion Model with Causal Generation and Cache Sharing (ICML 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
