Deep Double Descent via Smooth Interpolation.
Trans. Mach. Learn. Res., 2023
On the Lipschitz Constant of Deep Networks and Double Descent.
Proceedings of the 34th British Machine Vision Conference 2023, 2023
Are All Linear Regions Created Equal?
Proceedings of the International Conference on Artificial Intelligence and Statistics, 2022
Hyperplane Arrangements of Trained ConvNets Are Biased.
CoRR, 2020
On the Geometry of Rectifier Convolutional Neural Networks.
Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision Workshops, 2019