论文提出 x0-预测替代速度预测，改善高维潜空间扩散生成
On the Diffusibility of High-Dimensional Latents
如果你在调扩散模型，这篇讲清楚了为什么高维潜空间里速度预测费劲，换成 x0-预测就能稳定涨点，思路很实用。
来自 arXiv 的论文研究了表示自编码器（RAE）用于扩散模型时的问题：为图像重建微调预训练视觉编码器虽能恢复细节，却会降低表示的有效维度。论文指出在高维空间中，flow matching 的标准速度预测需要拟合信号流形之外的正交噪声方向，导致优化低效。改用干净数据参数化（x0-预测）后，在多个强重建编码器的文本到图像实验中一致提升生成表现。
On the Diffusibility of High-Dimensional Latents
Representation Autoencoders (RAEs) enable diffusion models to operate in the feature spaces of pretrained visual encoders. However, many off-the-shelf encoders are not optimized for faithful reconstruction, discarding fine-grained visual details. As expected, finetuning these encoders for image reconstruction recovers such details. However, perhaps counterintuitively, this procedure reduces the effective dimensionality of the resulting representation, and the altered geometry has downstream effects on generation. Specifically, we show that using the standard velocity prediction in flow matching in this high-dimensional space requires the model to fit orthogonal noise directions outside the low-dimensional signal manifold, making optimization inefficient. This motivates using the clean data parameterization ($\boldsymbol{x}_{0}$-prediction) instead, which focuses learning on the underlying signal manifold. Across experiments with multiple strong-reconstruction encoders, we show that $\boldsymbol{x}_{0}$-prediction consistently improves text-to-image generation performance.
