# MVAR: Visual Autoregressive Modeling with Scale and Spatial Markovian Conditioning (ICLR 2026)
[Jinhua Zhang](https://scholar.google.com/citations?user=tyYxiXoAAAAJ), [Wei Long](https://scholar.google.com/citations?user=CsVTBJoAAAAJ), [Minghao Han](https://scholar.google.com/citations?hl=en&user=IqrXj74AAAAJ), [Weiyi You](https://scholar.google.com/citations?user=q4uALoAAAAAJ), [Shuhang Gu](https://scholar.google.com/citations?user=-kSTt40AAAAJ)
[](https://arxiv.org/abs/2505.12742v3)
[](https://github.com/LabShuHangGU/MVAR)
[](https://nuanbaobao.github.io/MVAR)
[](https://huggingface.co/CVLUESTC/MVAR)