AI Paper Atlas

Research map / Video / Video generation

Video-GPT via Next Clip Diffusion

Shaobin Zhuang, Zhipeng Huang, Ying Zhang, Fangyikang Wang, Canmiao Fu, Binxin Yang, Chong Sun, Chen Li, Yali Wang

ICLR 2026 · Video › Video generation · approach: Compositional text-to-video diffusion · 8 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
increase frames per inference clip from 1 to 32 (ablation)Video-GPT · Physics-IQPhysics-IQ score: 0.00 -> 33.09Improved
tune history-conditioned CFG scale (3.0 vs 1) (ablation)Video-GPT · Physics-IQPhysics-IQ score: 25.09 -> 32.72Improved
raise CFG scale beyond 3.0 to 3.5 (ablation)Video-GPT · Physics-IQPhysics-IQ score: 32.72 -> 31.88Hurt
increase pretraining frames from 16 to 80 (ablation)Video-GPT · Physics-IQPhysics-IQ score: 22.06 -> 34.94Improved
add slight noise to clean clips during training (ablation)Video-GPT · Physics-IQPhysics-IQ score: 32.54 -> 33.09Improved
scale pretraining data from 1M to 70M videos (ablation)Video-GPT · Physics-IQPhysics-IQ score: 23.16 -> 33.09Improved
pretrain on video before fine-tuning on UCF-101 (ablation)Video-GPT · UCF-101 class-to-videoFVD: 489 -> 53Improved
next clip diffusion pretraining vs prior modelsVideo-GPT · Physics-IQPhysics-IQ score: 29.50 -> 34.97Improved

Earlier methods this paper compares with

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