English
Related papers

Related papers: Hidden Synergy: $L_1$ Weight Normalization and 1-P…

200 papers

Fast and precise Lipschitz constant estimation of neural networks is an important task for deep learning. Researchers have recently found an intrinsic trade-off between the accuracy and smoothness of neural networks, so training a network…

Machine Learning · Computer Science 2022-10-12 Zi Wang , Gautam Prakriya , Somesh Jha

In this work, we propose a new training method for finding minimum weight norm solutions in over-parameterized neural networks (NNs). This method seeks to improve training speed and generalization performance by framing NN training as a…

Machine Learning · Statistics 2018-06-22 Yamini Bansal , Madhu Advani , David D Cox , Andrew M Saxe

Neural networks often operate in the overparameterized regime, in which there are far more parameters than training samples, allowing the training data to be fit perfectly. That is, training the network effectively learns an interpolating…

Machine Learning · Computer Science 2025-03-19 Suzanna Parkinson , Greg Ongie , Rebecca Willett

Training neural networks under a strict Lipschitz constraint is useful for provable adversarial robustness, generalization bounds, interpretable gradients, and Wasserstein distance estimation. By the composition property of Lipschitz…

Machine Learning · Computer Science 2019-06-12 Cem Anil , James Lucas , Roger Grosse

This paper presents a margin-based multiclass generalization bound for neural networks that scales with their margin-normalized "spectral complexity": their Lipschitz constant, meaning the product of the spectral norms of the weight…

Machine Learning · Computer Science 2017-12-06 Peter Bartlett , Dylan J. Foster , Matus Telgarsky

We consider shallow (single hidden layer) neural networks and characterize their performance when trained with stochastic gradient descent as the number of hidden units $N$ and gradient descent steps grow to infinity. In particular, we…

Machine Learning · Statistics 2022-06-02 Jiahui Yu , Konstantinos Spiliopoulos

This paper presents a new bi-Lipschitz invertible neural network, the BiLipNet, which has the ability to smoothly control both its Lipschitzness (output sensitivity to input perturbations) and inverse Lipschitzness (input distinguishability…

Machine Learning · Computer Science 2024-06-07 Ruigang Wang , Krishnamurthy Dvijotham , Ian R. Manchester

We introduce Parseval networks, a form of deep neural networks in which the Lipschitz constant of linear, convolutional and aggregation layers is constrained to be smaller than 1. Parseval networks are empirically and theoretically…

Machine Learning · Statistics 2017-08-08 Moustapha Cisse , Piotr Bojanowski , Edouard Grave , Yann Dauphin , Nicolas Usunier

Using weight decay to penalize the L2 norms of weights in neural networks has been a standard training practice to regularize the complexity of networks. In this paper, we show that a family of regularizers, including weight decay, is…

Machine Learning · Computer Science 2022-06-09 Ziquan Liu , Yufei Cui , Antoni B. Chan

Batch Normalization (BN) has been proven to be quite effective at accelerating and improving the training of deep neural networks (DNNs). However, BN brings additional computation, consumes more memory and generally slows down the training…

Machine Learning · Computer Science 2019-05-23 Shuang Wu , Guoqi Li , Lei Deng , Liu Liu , Yuan Xie , Luping Shi

The channel estimation is one of important techniques to ensure reliable broadband signal transmission. Broadband channels are often modeled as a sparse channel. Comparing with traditional dense-assumption based linear channel estimation…

Information Theory · Computer Science 2014-07-24 Guan Gui , Li Xu , Fumiyuki Adachi

In this note, we study how neural networks with a single hidden layer and ReLU activation interpolate data drawn from a radially symmetric distribution with target labels 1 at the origin and 0 outside the unit ball, if no labels are known…

Machine Learning · Statistics 2022-09-05 Stephan Wojtowytsch

The Lipschitz constant of neural networks has been established as a key quantity to enforce the robustness to adversarial examples. In this paper, we tackle the problem of building $1$-Lipschitz Neural Networks. By studying Residual…

Machine Learning · Computer Science 2022-02-02 Laurent Meunier , Blaise Delattre , Alexandre Araujo , Alexandre Allauzen

Empirical studies have widely demonstrated that neural networks are highly sensitive to small, adversarial perturbations of the input. The worst-case robustness against these so-called adversarial examples can be quantified by the Lipschitz…

Machine Learning · Statistics 2025-07-03 Paul Geuchen , Dominik Stöger , Thomas Telaar , Felix Voigtlaender

In this paper, we propose a coupled tensor norm regularization that could enable the model output feature and the data input to lie in a low-dimensional manifold, which helps us to reduce overfitting. We show this regularization term is…

Optimization and Control · Mathematics 2023-02-24 Ying Gao , Yunfei Qu , Chunfeng Cui , Deren Han

We revisit the one-shot Neural Architecture Search (NAS) paradigm and analyze its advantages over existing NAS approaches. Existing one-shot method, however, is hard to train and not yet effective on large scale datasets like ImageNet. This…

Computer Vision and Pattern Recognition · Computer Science 2020-07-09 Zichao Guo , Xiangyu Zhang , Haoyuan Mu , Wen Heng , Zechun Liu , Yichen Wei , Jian Sun

While the class of Polynomial Nets demonstrates comparable performance to neural networks (NN), it currently has neither theoretical generalization characterization nor robustness guarantees. To this end, we derive new complexity bounds for…

Machine Learning · Computer Science 2022-02-11 Zhenyu Zhu , Fabian Latorre , Grigorios G Chrysos , Volkan Cevher

Modern neural networks are over-parametrized. In particular, each rectified linear hidden unit can be modified by a multiplicative factor by adjusting input and output weights, without changing the rest of the network. Inspired by the…

Computer Vision and Pattern Recognition · Computer Science 2019-02-28 Pierre Stock , Benjamin Graham , Rémi Gribonval , Hervé Jégou

We establish a layer-wise parameterization for 1D convolutional neural networks (CNNs) with built-in end-to-end robustness guarantees. In doing so, we use the Lipschitz constant of the input-output mapping characterized by a CNN as a…

Machine Learning · Computer Science 2024-01-26 Patricia Pauli , Ruigang Wang , Ian R. Manchester , Frank Allgöwer

We prove that, for the fundamental regression task of learning a single neuron, training a one-hidden layer ReLU network of any width by gradient flow from a small initialisation converges to zero loss and is implicitly biased to minimise…

Machine Learning · Computer Science 2023-10-03 Dmitry Chistikov , Matthias Englert , Ranko Lazic