AI Paper Atlas

Research map / Video / Video generation

Pathwise Test-Time Correction for Autoregressive Long Video Generation

Xunzhi Xiang, Zixuan Duan, Guiyu Zhang, Haiyu Zhang, Zhe Gao, Junta Wu, Shaofeng Zhang, Tengfei Wang, Qi Fan, Chunchao Guo

ECCV 2026 · Video › Video generation · approach: Autoregressive long video generation · 18 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
add pathwise test-time correctionSelf-Forcing · VBench 30-second generationSubject Consistency: 92.5 -> 94.0Improved
add pathwise test-time correctionCausVid · VBench 30-second generationSubject Consistency: 91.2 -> 93.2Improved
add pathwise test-time correctionCausVid · VBench 30-second generationDynamic Degree: 50.8 -> 69.5Improved
add pathwise test-time correctionSelf-Forcing · VBench 30-second generationDynamic Degree: 62.5 -> 60.2Hurt
add pathwise test-time correctionSelf-Forcing · 30-second color shiftColor-shift L1: 1.028 -> 0.644Improved
add pathwise test-time correctionSelf-Forcing · 30-second JEPAJEPA Difference: 0.191 -> 0.170Improved
add pathwise test-time correctionCausVid · VBench 30-second generationMotion Smoothness: 98.1 -> 97.6Hurt
use TTC instead of Best-of-N or Search-over-PathSelf-Forcing · 30-second test-time scalingSubject Consistency: 92.4/92.7 -> 94.0 at 10.53 vs 3.16 fpsImproved
use pathwise instead of single-point correction (ablation)TTC · Self-Forcing 30st-LPIPS boundary: 0.205 -> 0.176Improved
correct at timestep 750 only vs no correction (ablation)TTC · Self-Forcing 30st-LPIPS boundary: 0.178 -> 0.161Improved
correct at timestep 500 or 250 only vs no correction (ablation)TTC · Self-Forcing 30st-LPIPS boundary: 0.178 -> 0.182/0.183No clear effect
apply test-time optimization (LoRA) with reconstruction loss on early frames (ablation)distilled AR model · toy experiment (qualitative)drift and motionHurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 2 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