AI Paper Atlas

Research map / Video / Video generation

Self-Forcing++: Towards Minute-Scale High-Quality Video Generation

Justin Cui, Jie Wu, Ming Li, Tao Yang, Xiaojie Li, Rui Wang, Andrew Bai, Yuanhao Ban, Cho-Jui Hsieh

ICLR 2026 · Video › Video generation · approach: Efficient video diffusion models · 161 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
train with Self-Forcing++ (backward noise init, extended rollout) instead of Self-ForcingWan2.1-1.3B distilled AR video generator · 50s VBench Long + Gemini visual stabilityvisual stability: 40.12 -> 90.94Improved
Self-Forcing++ at 100s vs Self-Forcingsame · 100s dynamic degreedynamic degree: 26.41 -> 54.12Improved
Self-Forcing++ on short 5s clipssame · VBench 5stotal score: 83.00 -> 83.11No clear effect
shorten attention window in training to 9 latent frames (ablation)Self-Forcing · 50s visual stabilityvisual stability: 40.12 -> 52.50Mixed
add GRPO with optical-flow reward (ablation)Self-Forcing++ · long video temporal consistencyoptical-flow spikesImproved
scale training budget from 1x to 25x (ablation)Self-Forcing++ · long video generation lengthcoherent lengthImproved
autoregressive KV-cache generation (Self-Forcing++) vs overlapped recomputation methodsSelf-Forcing++ · temporal repetitionNoRepeat score: 67.19 (NOVA) / 73.44 (MAGI-1) vs 98.44Improved

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