English
Related papers

Related papers: Infinitely wide limits for deep Stable neural netw…

200 papers

In modern theoretical analyses of neural networks, the infinite-width limit is often invoked to justify Gaussian approximations of neuron preactivations (e.g., via neural network Gaussian processes or Tensor Programs). However, these…

Machine Learning · Computer Science 2026-01-05 Mana Sakai , Ryo Karakida , Masaaki Imaizumi

We study the extent to which wide neural networks may be approximated by Gaussian processes when initialized with random weights. It is a well-established fact that as the width of a network goes to infinity, its law converges to that of a…

Probability · Mathematics 2021-02-18 Ronen Eldan , Dan Mikulincer , Tselil Schramm

There has been a recent surge of interest in modeling neural networks (NNs) as Gaussian processes. In the limit of a NN of infinite width the NN becomes equivalent to a Gaussian process. Here we demonstrate that for an ensemble of large,…

Machine Learning · Computer Science 2019-08-28 Joseph M. Antognini

We prove a large deviation principle for deep neural networks with Gaussian weights and at most linearly growing activation functions, such as ReLU. This generalises earlier work, in which bounded and continuous activation functions were…

Machine Learning · Statistics 2026-02-10 Quirin Vogel

We consider fully connected and feedforward deep neural networks with dependent and possibly heavy-tailed weights, as introduced in [26], to address limitations of the standard Gaussian prior. It has been proved in [26] that, as the number…

Machine Learning · Statistics 2026-05-14 Nicola Apollonio , Giovanni Franzina , Giovanni Luca Torrisi

This work analyzes Graph Neural Networks, a generalization of Fully-Connected Deep Neural Nets on Graph structured data, when their width, that is the number of nodes in each fullyconnected layer is increasing to infinity. Infinite Width…

Machine Learning · Computer Science 2023-11-21 Yunus Cobanoglu

The gradient noise (GN) in the stochastic gradient descent (SGD) algorithm is often considered to be Gaussian in the large data regime by assuming that the \emph{classical} central limit theorem (CLT) kicks in. This assumption is often made…

Machine Learning · Statistics 2019-12-03 Umut Şimşekli , Mert Gürbüzbalaban , Thanh Huy Nguyen , Gaël Richard , Levent Sagun

Inducing and leveraging sparse activations during training and inference is a promising avenue for improving the computational efficiency of deep networks, which is increasingly important as network sizes continue to grow and their…

Machine Learning · Computer Science 2024-02-27 Ilan Price , Nicholas Daultry Ball , Samuel C. H. Lam , Adam C. Jones , Jared Tanner

A recent line of works studied wide deep neural networks (DNNs) by approximating them as Gaussian Processes (GPs). A DNN trained with gradient flow was shown to map to a GP governed by the Neural Tangent Kernel (NTK), whereas earlier works…

Machine Learning · Statistics 2021-12-15 Gadi Naveh , Oded Ben-David , Haim Sompolinsky , Zohar Ringel

A common theoretical approach to understanding neural networks is to take an infinite-width limit, at which point the outputs become Gaussian process (GP) distributed. This is known as a neural network Gaussian process (NNGP). However, the…

Machine Learning · Statistics 2025-06-26 Ben Anson , Edward Milsom , Laurence Aitchison

From the classical and influential works of Neal (1996), it is known that the infinite width scaling limit of a Bayesian neural network with one hidden layer is a Gaussian process, when the network weights have bounded prior variance.…

Machine Learning · Statistics 2024-06-06 Jorge Loría , Anindya Bhadra

Large width limits have been a recent focus of deep learning research: modulo computational practicalities, do wider networks outperform narrower ones? Answering this question has been challenging, as conventional networks gain…

Machine Learning · Computer Science 2021-11-09 Geoff Pleiss , John P. Cunningham

Contemporary deep neural networks exhibit impressive results on practical problems. These networks generalize well although their inherent capacity may extend significantly beyond the number of training examples. We analyze this behavior in…

Machine Learning · Computer Science 2015-09-03 Tamir Hazan , Tommi Jaakkola

The gradient noise (GN) in the stochastic gradient descent (SGD) algorithm is often considered to be Gaussian in the large data regime by assuming that the classical central limit theorem (CLT) kicks in. This assumption is often made for…

Machine Learning · Computer Science 2019-01-21 Umut Simsekli , Levent Sagun , Mert Gurbuzbalaban

In this paper, we characterize the noise of stochastic gradients and analyze the noise-induced dynamics during training deep neural networks by gradient-based optimizers. Specifically, we firstly show that the stochastic gradient noise…

Machine Learning · Computer Science 2021-09-22 Yixin Wu , Rui Luo , Chen Zhang , Jun Wang , Yaodong Yang

We propose stochastic, non-parametric activation functions that are fully learnable and individual to each neuron. Complexity and the risk of overfitting are controlled by placing a Gaussian process prior over these functions. The result is…

Machine Learning · Statistics 2017-12-01 Sebastian Urban , Marcus Basalla , Patrick van der Smagt

Neural network compression techniques have become increasingly popular as they can drastically reduce the storage and computation requirements for very large networks. Recent empirical studies have illustrated that even simple pruning…

Machine Learning · Statistics 2021-06-08 Melih Barsbey , Milad Sefidgaran , Murat A. Erdogdu , Gaël Richard , Umut Şimşekli

This paper is concerned with a special class of deep neural networks (DNNs) where the input and the output vectors have the same dimension. Such DNNs are widely used in applications, e.g., autoencoders. The training of such networks can be…

Machine Learning · Computer Science 2025-07-08 L. Berlyand , O. Krupchytskyi , V. Slavin

We study the Finite-Dimensional Distributions (FDDs) of deep neural networks with randomly initialized weights that have finite-order moments. Specifically, we establish Gaussian approximation bounds in the Wasserstein-$1$ norm between the…

Machine Learning · Statistics 2026-03-05 Krishnakumar Balasubramanian , Nathan Ross

It has long been known that a single-layer fully-connected neural network with an i.i.d. prior over its parameters is equivalent to a Gaussian process (GP), in the limit of infinite network width. This correspondence enables exact Bayesian…