The origins of generalization in neural nets are mysterious and have eluded understanding. We gain an intuitive grasp on generalization through carefully crafted experiments.
A pattern has emerged in which the majority of adversarial defenses are quickly broken by new attacks. Given the lack of success at generating robust defenses, we are led to ask a fundamental question: Are adversarial attacks inevitable?
Stacked U-Nets are simple, easy-to-train neural architecture for image segmentation and other image-to-image regression tasks. SUNets attain state of the art performance and fast inference with very few parameters.
Data poisoning is an adversarial attack in which examples are added to the training set of a classifier to manipulate the behavior of the model at test time. We propose a new poisoning attack that is effective on neural nets, and can be executed by an outsider with no control over the training process.
It is well known that certain neural network architectures produce loss functions that train easier and generalize better, but the reasons for this are not well understood. To understand this better, we explore the structure of neural loss functions using a range of visualization methods.