AI Paper Atlas

Research map / Video / Video generation

Context Forcing: Consistent Autoregressive Video Generation with Long Context

Shuo Chen, Cong Wei, Sun Sun, Tiancheng SHEN, Ping Nie, Kai Zou, Ge Zhang, Ming-Hsuan Yang, Wenhu Chen

ICML 2026 · Video › Video generation · approach: Autoregressive long video generation · 0 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 contextual distribution matching distillation (ablation)Context Forcing · VBench 60s videototal score: 80.36 -> 82.45Improved
add contextual distribution matching distillation (ablation)Context Forcing · VBench 60s videosemantic score: 72.70 -> 76.10Improved
add contextual distribution matching distillation (ablation)Context Forcing · VBench 60s videodynamic degree: 48.12 -> 58.26Improved
add bounded positional encoding (ablation)Context Forcing · VBench 60s videototal score: 73.52 -> 82.45Improved
add bounded positional encoding (ablation)Context Forcing · VBench 60s videosubject consistency: 79.24 -> 94.88Improved
add bounded positional encoding (ablation)Context Forcing · VBench 60s videobackground consistency: 84.68 -> 95.34Improved
use similarity-based slow memory selection instead of uniform interval 1 (ablation)Context Forcing · VBench 60s videototal score: 80.82 -> 82.45Improved
use similarity-based slow memory selection instead of uniform interval 2 (ablation)Context Forcing · VBench 60s videototal score: 81.11 -> 82.45Improved
use similarity-based slow memory selection instead of uniform interval 1 (ablation)Context Forcing · VBench 60s videosemantic score: 75.32 -> 76.10No clear effect
Context Forcing student versus LongLive on 60s DINOContext Forcing · Single-prompt 60s consistencyDINO score at 60s: 86.26 -> 87.89Improved
Context Forcing student versus LongLive on 60s VBench totalContext Forcing · VBench 60s videototal score: 83.64 -> 82.45Hurt
Context Forcing student versus LongLive on 5s VBench totalContext Forcing · VBench 5s videototal score: 84.87 -> 83.44Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 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