Optimal Initialization in Depth: Lyapunov Initialization and Limit Theorems for Deep Leaky ReLU Networks

• URL: http://arxiv.org/abs/2602.10949v1
• Date: Wed, 11 Feb 2026 15:36:13 GMT
• Title: Optimal Initialization in Depth: Lyapunov Initialization and Limit Theorems for Deep Leaky ReLU Networks
• Authors: Constantin Kogler, Tassilo Schwarz, Samuel Kittle,
• Abstract summary: We show that as the number of layers increases, their growth is governed by a parameter called the Lyapunov exponent.<n>We propose a novel method for setting the Lyapunov exponent to zero, which ensures that the neural network is as stable as possible.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The development of effective initialization methods requires an understanding of random neural networks. In this work, a rigorous probabilistic analysis of deep unbiased Leaky ReLU networks is provided. We prove a Law of Large Numbers and a Central Limit Theorem for the logarithm of the norm of network activations, establishing that, as the number of layers increases, their growth is governed by a parameter called the Lyapunov exponent. This parameter characterizes a sharp phase transition between vanishing and exploding activations, and we calculate the Lyapunov exponent explicitly for Gaussian or orthogonal weight matrices. Our results reveal that standard methods, such as He initialization or orthogonal initialization, do not guarantee activation stabilty for deep networks of low width. Based on these theoretical insights, we propose a novel initialization method, referred to as Lyapunov initialization, which sets the Lyapunov exponent to zero and thereby ensures that the neural network is as stable as possible, leading empirically to improved learning.

Related papers

• Quantization vs Pruning: Insights from the Strong Lottery Ticket Hypothesis [5.494111035517599]
  Quantization is an essential technique for making neural networks more efficient, yet our theoretical understanding of it remains limited.<n>Previous works demonstrated that extremely low-precision networks, such as binary networks, can be constructed by pruning large, randomly- approximationd networks.<n>We build on foundational results by Borgs et al. on the Number Partitioning Problem to derive new theoretical results for the Random Subset Sum Problem in a quantized setting.
  arXiv  Detail & Related papers  (2025-08-14T18:51:34Z)
• Globally Optimal Training of Neural Networks with Threshold Activation Functions [63.03759813952481]
  We study weight decay regularized training problems of deep neural networks with threshold activations. We derive a simplified convex optimization formulation when the dataset can be shattered at a certain layer of the network.
  arXiv  Detail & Related papers  (2023-03-06T18:59:13Z)
• Neural Networks with Sparse Activation Induced by Large Bias: Tighter Analysis with Bias-Generalized NTK [86.45209429863858]
  We study training one-hidden-layer ReLU networks in the neural tangent kernel (NTK) regime. We show that the neural networks possess a different limiting kernel which we call textitbias-generalized NTK We also study various properties of the neural networks with this new kernel.
  arXiv  Detail & Related papers  (2023-01-01T02:11:39Z)
• Subquadratic Overparameterization for Shallow Neural Networks [60.721751363271146]
  We provide an analytical framework that allows us to adopt standard neural training strategies. We achieve the desiderata viaak-Lojasiewicz, smoothness, and standard assumptions.
  arXiv  Detail & Related papers  (2021-11-02T20:24:01Z)
• A Weight Initialization Based on the Linear Product Structure for Neural Networks [0.0]
  We study neural networks from a nonlinear point of view and propose a novel weight initialization strategy that is based on the linear product structure (LPS) of neural networks. The proposed strategy is derived from the approximation of activation functions by using theories of numerical algebra to guarantee to find all the local minima.
  arXiv  Detail & Related papers  (2021-09-01T00:18:59Z)
• On the Explicit Role of Initialization on the Convergence and Implicit Bias of Overparametrized Linear Networks [1.0323063834827415]
  We present a novel analysis of single-hidden-layer linear networks trained under gradient flow. We show that the squared loss converges exponentially to its optimum. We derive a novel non-asymptotic upper-bound on the distance between the trained network and the min-norm solution.
  arXiv  Detail & Related papers  (2021-05-13T15:13:51Z)
• A Convergence Theory Towards Practical Over-parameterized Deep Neural Networks [56.084798078072396]
  We take a step towards closing the gap between theory and practice by significantly improving the known theoretical bounds on both the network width and the convergence time. We show that convergence to a global minimum is guaranteed for networks with quadratic widths in the sample size and linear in their depth at a time logarithmic in both. Our analysis and convergence bounds are derived via the construction of a surrogate network with fixed activation patterns that can be transformed at any time to an equivalent ReLU network of a reasonable size.
  arXiv  Detail & Related papers  (2021-01-12T00:40:45Z)
• Fractional moment-preserving initialization schemes for training deep neural networks [1.14219428942199]
  A traditional approach to deep neural networks (DNNs) is to sample the network weights randomly for preserving the variance of pre-activations. In this paper, we show that weights and therefore pre-activations can be modeled with a heavy-tailed distribution. We show through numerical experiments that our schemes can improve the training and test performance.
  arXiv  Detail & Related papers  (2020-05-25T01:10:01Z)
• Revisiting Initialization of Neural Networks [72.24615341588846]
  We propose a rigorous estimation of the global curvature of weights across layers by approximating and controlling the norm of their Hessian matrix. Our experiments on Word2Vec and the MNIST/CIFAR image classification tasks confirm that tracking the Hessian norm is a useful diagnostic tool.
  arXiv  Detail & Related papers  (2020-04-20T18:12:56Z)
• MSE-Optimal Neural Network Initialization via Layer Fusion [68.72356718879428]
  Deep neural networks achieve state-of-the-art performance for a range of classification and inference tasks. The use of gradient combined nonvolutionity renders learning susceptible to novel problems. We propose fusing neighboring layers of deeper networks that are trained with random variables.
  arXiv  Detail & Related papers  (2020-01-28T18:25:15Z)
• Provable Benefit of Orthogonal Initialization in Optimizing Deep Linear Networks [39.856439772974454]
  We show that the width needed for efficient convergence to a global minimum is independent of the depth. Our results suggest an explanation for the recent empirical successes found by initializing very deep non-linear networks according to the principle of dynamical isometry.
  arXiv  Detail & Related papers  (2020-01-16T18:48:34Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.