AI Paper Atlas

Research map / Video / Video generation

UltraViCo: Breaking Extrapolation Limits in Video Diffusion Transformers

Min Zhao, Hongzhou Zhu, Yingze Wang, Bokai Yan, Jintao Zhang, Guande He, Ling Yang, Chongxuan Li, Jun Zhu

ICLR 2026 · Video › Video generation · approach: Efficient video diffusion models · 11 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
suppress attention to out-of-window tokens (UltraViCo) vs best baseline RIFLExHunyuanVideo 4x · VBench 4x extrapolationImaging Quality: 50.01 -> 66.54Improved
UltraViCo vs prior methods on Wan beyond 3xWan2.1-1.3B · VBench 4x extrapolationDynamic Degree: 11 -> 47Improved
UltraViCo vs PE on content repetitionHunyuanVideo 4x · VBench 4x extrapolationNoRepeat Score: 31.41 -> 99.87Improved
UltraViCo on overall consistency vs best baseline at 4xHunyuanVideo 4x · VBench 4x extrapolationConsistency: 0.9915 -> 0.9468Hurt
mask tokens at harmonic alignment positions (ablation)HunyuanVideo · periodic attentionrepetition and qualityImproved
increase masking ratio of out-of-window attention (ablation)video DiT · controlled experimentvideo qualityImproved
use parabolic or linear decay instead of constant decay (ablation)UltraViCo · Wan 3x extrapolationqualityNo clear effect
lower decay factor alpha below 0.9 (ablation)UltraViCo · Wan 3x extrapolationconsistencyHurt
lower beta below 0.6 at harmonic positions (ablation)UltraViCo · HunyuanVideoconsistencyHurt
add UltraViCo to sliding window (ablation)Sliding Window · Wan 6x extrapolationConsistency: 0.8478 -> 0.9183Improved
add UltraViCo to FreeNoise (ablation)FreeNoise · Wan 6x extrapolationConsistency: 0.9243 -> 0.9431Improved
add UltraViCo to FIFO-Diffusion (ablation)FIFO-Diffusion · Wan 6x extrapolationConsistency: 0.9131 -> 0.9319Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

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

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