We present Visual AutoRegressive modeling (VAR), a new generation paradigm that redefines the autoregressive learning on images as coarse-to-fine "next-scale prediction" or "next-resolution prediction", diverging from the standard raster-scan "next-token prediction". This simple, intuitive methodology allows autoregressive (AR) transformers to learn visual distributions fast and generalize well: VAR, for the first time, makes AR models surpass diffusion transformers in image generation. On ImageNet 256x256 benchmark, VAR significantly improve AR baseline by improving Frechet inception distance (FID) from 18.65 to 1.80, inception score (IS) from 80.4 to 356.4, with around 20x faster inference speed. It is also empirically verified that VAR outperforms the Diffusion Transformer (DiT) in multiple dimensions including image quality, inference speed, data efficiency, and scalability. Scaling up VAR models exhibits clear power-law scaling laws similar to those observed in LLMs, with linear correlation coefficients near -0.998 as solid evidence. VAR further showcases zero-shot generalization ability in downstream tasks including image in-painting, out-painting, and editing. These results suggest VAR has initially emulated the two important properties of LLMs: Scaling Laws and zero-shot task generalization. We have released all models and codes to promote the exploration of AR/VAR models for visual generation and unified learning.

翻译：我们提出视觉自回归建模（VAR），这是一种新的生成范式，将图像上的自回归学习重新定义为由粗到精的“下一尺度预测”或“下一分辨率预测”，与标准的栅格扫描式“下一标记预测”不同。这种简单直观的方法使自回归（AR）变换器能够快速学习视觉分布并具有良好的泛化能力：VAR首次使得AR模型在图像生成方面超越扩散变换器。在ImageNet 256x256基准测试中，VAR通过将Fréchet初始距离（FID）从18.65提升至1.80，初始分数（IS）从80.4提升至356.4，显著改进了AR基线，同时推理速度提升约20倍。实证还验证了VAR在图像质量、推理速度、数据效率和可扩展性等多个维度上优于扩散变换器（DiT）。扩展VAR模型展现出与大型语言模型相似清晰的幂律缩放法则，线性相关系数接近-0.998，这是有力的证据。VAR还进一步展示了在下游任务（包括图像修复、外推和编辑）中的零样本泛化能力。这些结果表明VAR初步模拟了大型语言模型的两个重要特性：缩放法则和零样本任务泛化。我们已发布所有模型和代码，以促进AR/VAR模型在视觉生成和统一学习中的探索。