Awesome World Models / arXiv

Self Forcing: Bridging the Train-Test Gap in Autoregressive Video Diffusion

作者：作者信息待提取 · 单位：The University of Texas at Austin · 来源：General Approaches to World Models · 发布日期：2025 · 分类：2. Building World Models from 2D Vision Priors / Useful Approaches in Video Generation

二维视觉先验视频生成
新标签打开 AlphaXiv新标签打开 Project新标签打开 Code
Self Forcing: Bridging the Train-Test Gap in Autoregressive Video Diffusion figure
AlphaXiv 中文概览（可滚动查看）