English
Related papers

Related papers: Beyond Gaussian Initializations: Signal Preserving…

200 papers

There is a growing literature on the study of large-width properties of deep Gaussian neural networks (NNs), i.e. deep NNs with Gaussian-distributed parameters or weights, and Gaussian stochastic processes. Motivated by some empirical and…

Machine Learning · Computer Science 2023-04-11 Alberto Bordino , Stefano Favaro , Sandra Fortini

We establish novel rates for the Gaussian approximation of random deep neural networks with Gaussian parameters (weights and biases) and Lipschitz activation functions, in the wide limit. Our bounds apply for the joint output of a network…

Statistics Theory · Mathematics 2023-12-20 Dario Trevisan

Weight initialization is important for faster convergence and stability of deep neural networks training. In this paper, a robust initialization method is developed to address the training instability in long short-term memory (LSTM)…

Appropriate weight initialization settings, along with the ReLU activation function, have become cornerstones of modern deep learning, enabling the training and deployment of highly effective and efficient neural network models across…

Machine Learning · Computer Science 2024-04-02 Hyunwoo Lee , Yunho Kim , Seung Yeop Yang , Hayoung Choi

A proper initialization of the weights in a neural network is critical to its convergence. Current insights into weight initialization come primarily from linear activation functions. In this paper, I develop a theory for weight…

Machine Learning · Computer Science 2017-05-04 Siddharth Krishna Kumar

We prove large and moderate deviations for the output of Gaussian fully connected neural networks. The main achievements concern deep neural networks (i.e., when the model has more than one hidden layer) and hold for bounded and continuous…

Probability · Mathematics 2026-04-01 Claudio Macci , Barbara Pacchiarotti , Giovanni Luca Torrisi

We study the problem of training deep neural networks with Rectified Linear Unit (ReLU) activation function using gradient descent and stochastic gradient descent. In particular, we study the binary classification problem and show that for…

Machine Learning · Computer Science 2018-12-31 Difan Zou , Yuan Cao , Dongruo Zhou , Quanquan Gu

Overparameterized models may have many interpolating solutions; implicit regularization refers to the hidden preference of a particular optimization method towards a certain interpolating solution among the many. A by now established line…

Machine Learning · Computer Science 2024-09-18 Hung-Hsu Chou , Holger Rauhut , Rachel Ward

Stochastic variational inference is an established way to carry out approximate Bayesian inference for deep models. While there have been effective proposals for good initializations for loss minimization in deep learning, far less…

Machine Learning · Statistics 2019-01-28 Simone Rossi , Pietro Michiardi , Maurizio Filippone

To theoretically understand the behavior of trained deep neural networks, it is necessary to study the dynamics induced by gradient methods from a random initialization. However, the nonlinear and compositional structure of these models…

Machine Learning · Computer Science 2021-12-21 Karl Hajjar , Lénaïc Chizat , Christophe Giraud

Weight initialization significantly impacts the convergence and performance of neural networks. While traditional methods like Xavier and Kaiming initialization are widely used, they often fall short for spiking neural networks (SNNs),…

Machine Learning · Computer Science 2024-11-28 Da Chang , Deliang Wang , Xiao Yang

An activation function has crucial role in a deep neural network. A simple rectified linear unit (ReLU) are widely used for the activation function. In this paper, a weighted sigmoid gate unit (WiG) is proposed as the activation function.…

Computer Vision and Pattern Recognition · Computer Science 2018-10-04 Masayuki Tanaka

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

Gaussian splatting (GS) along with its extensions and variants provides outstanding performance in real-time scene rendering while meeting reduced storage demands and computational efficiency. While the selection of 2D images capturing the…

Computer Vision and Pattern Recognition · Computer Science 2025-12-04 Konstantinos D. Polyzos , Athanasios Bacharis , Saketh Madhuvarasu , Nikos Papanikolopoulos , Tara Javidi

It has been known for a long time that initializing weight matrices to be orthogonal instead of having i.i.d. Gaussian components can improve training performance. This phenomenon can be analyzed using finite-width corrections, where the…

Machine Learning · Computer Science 2026-05-15 Max Guillen , Jan E. Gerken

In this paper, we study the infinite-depth limit of finite-width residual neural networks with random Gaussian weights. With proper scaling, we show that by fixing the width and taking the depth to infinity, the pre-activations converge in…

Machine Learning · Statistics 2023-01-16 Soufiane Hayou

Weight initialization governs signal propagation and gradient flow at the start of training. This paper offers a theory-grounded and empirically validated study across two regimes: compact ReLU multilayer perceptrons and GPT-2-style…

Machine Learning · Computer Science 2025-10-13 Yankun Han

Proper weight initialization prior to training has historically been one of the key factors that helped kick off the deep learning revolution. Initialization is even more crucial in "reservoir computing", where the weights of a readout…

Machine Learning · Computer Science 2026-05-12 Tommaso Fioratti , Riccardo Marcaccioli , Francesco Casola

The approximation and convergence properties of implicit neural representations (INRs) are known to be highly sensitive to parameter initialization strategies. While several data-driven initialization methods demonstrate significant…

Computer Vision and Pattern Recognition · Computer Science 2026-04-01 Kushal Vyas , Alper Kayabasi , Daniel Kim , Vishwanath Saragadam , Ashok Veeraraghavan , Guha Balakrishnan

We identify and address a fundamental limitation of sinusoidal representation networks (SIRENs), a class of implicit neural representations. SIRENs Sitzmann et al. (2020), when not initialized appropriately, can struggle at fitting signals…

Computer Vision and Pattern Recognition · Computer Science 2025-09-17 Hemanth Chandravamsi , Dhanush V. Shenoy , Steven H. Frankel