TL;DR该论文介绍了Bayes by Hypernet，一种新的变分逼近方法，通过将超网络视为隐式分布来解决现代神经网络在未见过的、嘈杂的或标记错误的数据上过于自信，并且不能产生有意义的不确定性度量的短板，本文在MNIST和CIFAR5任务中表现优异且最具鲁棒性，同时满足复杂度、可扩展性和准确度的要求。
Abstract
We interpret hypernetworks within the framework of variational inference within implicit distributions. Our method, Bayes by Hypernet, is able to model a richer variational distribution than previous methods. Experiments show that it achieves comparable predictive performance on the MN