AI Paper Atlas

Research map / Video / Video generation

Causality in Video Diffusers is Separable from Denoising

Xingjian Bai, Guande He, Zhengqi Li, Eli Shechtman, Xun Huang, Zongze Wu

CVPR 2026 · Video › Video generation · approach: Efficient video diffusion models · 5 citations (OpenAlex)

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 changedWhereMeasuredEffect
decouple causal encoder from frame-wise diffusion decoder (SCD-M vs Causal DiT-M)causal DiT · TECO-MinecraftFVD / Sec per frame: FVD 38.7 -> 37.6; 2.4 -> 0.52 s/frameImproved
decouple causal encoder from decoder (SCD-B vs Causal DiT-B)causal DiT · UCF-101FVD / Sec per frame: FVD 187.6 -> 174.7; 3.9 -> 1.1 s/frameImproved
increase causal encoder depth (ablation)SCD-M · TECO-MinecraftFVD: 37.6 -> 36.1Improved
increase causal encoder depth (ablation)SCD-B · UCF-101FVD: 174.7 -> 171.1Improved
increase diffusion decoder depth (ablation)SCD-M · TECO-MinecraftFVD / Sec per frame: FVD 37.6 -> 34.9; 0.52 -> 1.6 s/frameMixed
increase diffusion decoder depth (ablation)SCD-B · UCF-101FVD / Sec per frame: FVD 174.7 -> 158.7; 1.1 -> 2.8 s/frameMixed
replace Self Forcing with SCD fine-tuned from Wan2.1 1.3BSelf Forcing · VBenchthroughput (FPS): 8.9 -> 11.1 FPS; total score 84.26 -> 84.03Mixed
skip 15 middle layers across later denoising steps then finetune (ablation)AR Wan2.1 1.3B · qualitative videosvisual fidelityNo clear effect
switch last 5 layers to frame-diagonal attention with 5K-step finetune (ablation)AR Wan2.1 1.3B · qualitative videosvisual fidelityNo clear effect
use frame-wise concatenation instead of channel-wise concatenation of context and noisy frame (ablation)SCD decoder · SCD pretraining ablationqualityImproved
inject Gaussian noise into causal context (ablation)SCD · SCD pretraining ablationqualityImproved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy