Why neural networks find simple solutions: the many regularizers of
geometric complexity

Why neural networks find simple solutions: the many regularizers of geometric complexity

27 September 2022

Papers citing "Why neural networks find simple solutions: the many regularizers of geometric complexity"

10 / 10 papers shown

Title
A Margin-based Multiclass Generalization Bound via Geometric Complexity Michael Munn Benoit Dherin Javier Gonzalvo UQCV 27 2 0 28 May 2024
From Robustness to Improved Generalization and Calibration in Pre-trained Language Models Josip Jukić Jan Snajder 21 0 0 31 Mar 2024
Neural Redshift: Random Networks are not Random Functions Damien Teney A. Nicolicioiu Valentin Hartmann Ehsan Abbasnejad 86 18 0 04 Mar 2024
Why Does Little Robustness Help? Understanding and Improving Adversarial Transferability from Surrogate Training Yechao Zhang Shengshan Hu Leo Yu Zhang Junyu Shi Minghui Li Xiaogeng Liu Wei Wan Hai Jin AAML 22 20 0 15 Jul 2023
Mathematical Challenges in Deep Learning V. Nia Guojun Zhang I. Kobyzev Michael R. Metel Xinlin Li ... S. Hemati M. Asgharian Linglong Kong Wulong Liu Boxing Chen AI4CE VLM 22 1 0 24 Mar 2023
On the Lipschitz Constant of Deep Networks and Double Descent Matteo Gamba Hossein Azizpour Marten Bjorkman 16 6 0 28 Jan 2023
Deep Double Descent via Smooth Interpolation Matteo Gamba Erik Englesson Marten Bjorkman Hossein Azizpour 51 10 0 21 Sep 2022
Stochastic Training is Not Necessary for Generalization Jonas Geiping Micah Goldblum Phillip E. Pope Michael Moeller Tom Goldstein 81 72 0 29 Sep 2021
Geometric deep learning: going beyond Euclidean data M. Bronstein Joan Bruna Yann LeCun Arthur Szlam P. Vandergheynst GNN 231 3,202 0 24 Nov 2016
On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima N. Keskar Dheevatsa Mudigere J. Nocedal M. Smelyanskiy P. T. P. Tang ODL 273 2,696 0 15 Sep 2016