English
Related papers

Related papers: Neural Collapse Dynamics: Depth, Activation, Regul…

200 papers

Deep Neural Networks (DNNs) share important similarities with structural glasses. Both have many degrees of freedom, and their dynamics are governed by a high-dimensional, non-convex landscape representing either the loss or energy,…

Computational Physics · Physics 2025-03-25 Max Kerr Winter , Liesbeth M. C. Janssen

Understanding the dynamics of neural networks in different width regimes is crucial for improving their training and performance. We present an exact solution for the learning dynamics of a one-hidden-layer linear network, with…

Machine Learning · Computer Science 2025-02-24 Yizhou Xu , Liu Ziyin

We identify and formalize a fundamental gradient descent phenomenon resulting in a learning proclivity in over-parameterized neural networks. Gradient Starvation arises when cross-entropy loss is minimized by capturing only a subset of…

Machine Learning · Computer Science 2021-11-25 Mohammad Pezeshki , Sékou-Oumar Kaba , Yoshua Bengio , Aaron Courville , Doina Precup , Guillaume Lajoie

Deep Learning has revolutionized vision via convolutional neural networks (CNNs) and natural language processing via recurrent neural networks (RNNs). However, success stories of Deep Learning with standard feed-forward neural networks…

Machine Learning · Computer Science 2018-01-31 Günter Klambauer , Thomas Unterthiner , Andreas Mayr , Sepp Hochreiter

Recent experiments in computer vision demonstrate texture bias as the primary reason for supreme results in models employing Convolutional Neural Networks (CNNs), conflicting with early works claiming that these networks identify objects…

Computer Vision and Pattern Recognition · Computer Science 2023-08-29 Satyam Mohla , Anshul Nasery , Biplab Banerjee

In self-supervised representation learning, a common idea behind most of the state-of-the-art approaches is to enforce the robustness of the representations to predefined augmentations. A potential issue of this idea is the existence of…

Machine Learning · Computer Science 2021-08-26 Tianyu Hua , Wenxiao Wang , Zihui Xue , Sucheng Ren , Yue Wang , Hang Zhao

In order to develop complex relationships between their inputs and outputs, deep neural networks train and adjust large number of parameters. To make these networks work at high accuracy, vast amounts of data are needed. Sometimes, however,…

Machine Learning · Computer Science 2022-01-19 Joshua Shunk

Network robustness is critical for various societal and industrial networks again malicious attacks. In particular, connectivity robustness and controllability robustness reflect how well a networked system can maintain its connectedness…

Systems and Control · Electrical Eng. & Systems 2023-07-25 Yang Lou , Ruizi Wu , Junli Li , Lin Wang , Xiang Li , Guanrong Chen

Deep neural networks are renowned for their ability to generalise well across diverse tasks, even when heavily overparameterized. Existing works offer only partial explanations (for example, the NTK-based task-model alignment explanation…

Machine Learning · Computer Science 2025-06-02 Chris Mingard , Lukas Seier , Niclas Göring , Andrei-Vlad Badelita , Charles London , Ard Louis

Recurrent Neural Networks (RNNs) frequently exhibit complicated dynamics, and their sensitivity to the initialization process often renders them notoriously hard to train. Recent works have shed light on such phenomena analyzing when…

Machine Learning · Computer Science 2022-10-12 Vaggos Chatziafratis , Ioannis Panageas , Clayton Sanford , Stelios Andrew Stavroulakis

We consider the variational problem of cross-entropy loss with $n$ feature vectors on a unit hypersphere in $\mathbb{R}^d$. We prove that when $d \geq n - 1$, the global minimum is given by the simplex equiangular tight frame, which…

Machine Learning · Computer Science 2021-01-20 Jianfeng Lu , Stefan Steinerberger

We analyze the dynamics of finite width effects in wide but finite feature learning neural networks. Starting from a dynamical mean field theory description of infinite width deep neural network kernel and prediction dynamics, we provide a…

Machine Learning · Statistics 2023-11-08 Blake Bordelon , Cengiz Pehlevan

Substantial work indicates that the dynamics of neural networks (NNs) is closely related to their initialization of parameters. Inspired by the phase diagram for two-layer ReLU NNs with infinite width (Luo et al., 2021), we make a step…

Machine Learning · Computer Science 2022-10-20 Hanxu Zhou , Qixuan Zhou , Zhenyuan Jin , Tao Luo , Yaoyu Zhang , Zhi-Qin John Xu

Recent efforts have shown machine learning to be useful for the prediction of nonlinear fluid dynamics. Predictive accuracy is often a central motivation for employing neural networks, but the pattern recognition central to the network…

Fluid Dynamics · Physics 2022-08-23 Shizheng Wen , Michael W. Lee , Kai M. Kruger Bastos , Earl H. Dowell

Catastrophic forgetting (CF) happens whenever a neural network overwrites past knowledge while being trained on new tasks. Common techniques to handle CF include regularization of the weights (using, e.g., their importance on past tasks),…

Machine Learning · Statistics 2021-08-06 Jary Pomponi , Simone Scardapane , Aurelio Uncini

We study the effect of width on the dynamics of feature-learning neural networks across a variety of architectures and datasets. Early in training, wide neural networks trained on online data have not only identical loss curves but also…

Machine Learning · Computer Science 2023-12-07 Nikhil Vyas , Alexander Atanasov , Blake Bordelon , Depen Morwani , Sabarish Sainathan , Cengiz Pehlevan

Training neural networks with batch normalization and weight decay has become a common practice in recent years. In this work, we show that their combined use may result in a surprising periodic behavior of optimization dynamics: the…

Machine Learning · Computer Science 2022-01-19 Ekaterina Lobacheva , Maxim Kodryan , Nadezhda Chirkova , Andrey Malinin , Dmitry Vetrov

We introduce the nested canalyzing depth of a function, which measures the extent to which it retains a nested canalyzing structure. We characterize the structure of functions with a given depth and compute the expected activities and…

Molecular Networks · Quantitative Biology 2012-03-01 Lori Layne , Elena Dimitrova , Matthew Macauley

Convolutional neural networks (CNNs) can model complicated non-linear relations between images. However, they are notoriously sensitive to small changes in the input. Most CNNs trained to describe image-to-image mappings generate temporally…

Computer Vision and Pattern Recognition · Computer Science 2020-04-15 Gabriel Eilertsen , Rafał K. Mantiuk , Jonas Unger

When training deep neural networks with gradient descent, sharpness often increases -- a phenomenon known as progressive sharpening -- before saturating at the edge of stability. Although commonly observed in practice, the underlying…

Machine Learning · Computer Science 2025-06-10 Geonhui Yoo , Minhak Song , Chulhee Yun