Zhaoyuan Yang

Dropout Inference with Non-Uniform Weight Scaling

2022 · arXiv (Cornell University)

Dropout as regularization has been used extensively to prevent overfitting for training neural networks. During training, units and their connections are randomly dropped, which could be considered as sampling many different submodels from the original …

Adversarial Purification with the Manifold Hypothesis

2024 · Proceedings of the AAAI Conference on Artificial Intelligence

In this work, we formulate a novel framework for adversarial robustness using the manifold hypothesis. This framework provides sufficient conditions for defending against adversarial examples. We develop an adversarial purification method with this framework. Our …