Several works in implicit and explicit generative modeling empirically observed that feature-learning discriminators outperform fixed-kernel discriminators in terms of the sample quality of the models. We provide separation results between probability metrics with fixed-kernel and feature-learning discriminators using the function classes $\mathcal{F}_2$ and $\mathcal{F}_1$ respectively, which were developed to study overparametrized two-layer neural networks. In particular, we construct pairs of distributions over hyper-spheres that can not be discriminated by fixed kernel $(\mathcal{F}_2)$ integral probability metric (IPM) and Stein discrepancy (SD) in high dimensions, but that can be discriminated by their feature learning ($\mathcal{F}_1$) counterparts. To further study the separation we provide links between the $\mathcal{F}_1$ and $\mathcal{F}_2$ IPMs with sliced Wasserstein distances. Our work suggests that fixed-kernel discriminators perform worse than their feature learning counterparts because their corresponding metrics are weaker.
翻译:在隐含和明显的基因模型中,有几项工作是隐含的和明显的基因模型,从模型的样本质量来看,特征学习歧视者比固定内核歧视者要优于固定内核歧视者。我们提供使用功能类的固定内核和特征学习歧视者概率指标的分离结果,使用功能类分别为$\mathcal{F ⁇ 2$和$\mathcal{F ⁇ 1$}F ⁇ 1$。开发这些参数是为了研究过度平衡的双层神经网络。特别是,我们建造了超高外层的分布配对,不能被固定内核(mathcal{F ⁇ 2)$(mathcal{F ⁇ 2$) 整体概率指标(IPM) 和 Stein 差异(SD) 高维度差异(SD),但可能因其特征学习结果而有所区别。为了进一步研究我们提供的分离联系是$\mathcal{F ⁇ 1$和 $\\cathcal{F ⁇ 2$IPMMPMs 距离切片。我们的工作表明,固定内层歧视者的表现比其特征学习对应指标要差得多,因为它们对应指标较弱。