Spectral phase transitions and trainability in neural network learning dynamics

• URL: http://arxiv.org/abs/2606.28486v1
• Date: Fri, 26 Jun 2026 18:00:01 GMT
• Title: Spectral phase transitions and trainability in neural network learning dynamics
• Abstract summary: We study the emergence of low-dimensional structures in the spectra of neural network weight matrices during training.<n>Our results suggest that spectral analysis may provide a unified perspective of learning dynamics.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The emergence of low-dimensional structures in the spectra of neural network weight matrices is a common empirical feature of trained models, but the dynamical origin of this phenomenon during learning remains an open problem. We formulate neural network training as the stochastic evolution of an initially random matrix ensemble, driven by stochastic gradient descent (SGD) updates that reshape the spectral bulk while amplifying signal strength. This induces a Baik-Ben Arous-Péché (BBP) transition during training, where isolated eigenvalues detach from the random bulk distribution, providing a dynamical framework for representation formation in high-dimensional learning dynamics. We demonstrate this in a solvable linear teacher-student model, where spectral evolution is analytically tractable and a phase diagram of trainability governed by the step size (or learning rate) and initial weight variance is obtained, and subsequently extend our formalism beyond the linear regime to nonlinear and stochastic settings. Numerical simulations in realistic settings support this picture, showing robust emergence of spectral alignment during training. Our results suggest that spectral analysis may provide a unified perspective of stochastic learning dynamics, linking trainability, optimisation hyperparameters, spectral phase transitions, and representation learning in neural networks.

Related papers

• The Weight Gram Matrix Captures Sequential Feature Linearization in Deep Networks [10.475180266869081]
  Understanding how deep neural networks learn remains a central challenge in machine learning theory.<n>We propose a feature-centric framework for analyzing neural network training by relating weight updates to feature evolution.<n>By analyzing the training and layer-wise dynamics, we show that deep networks learn to sequentially transform representations toward target-linear structure.
  arXiv  Detail & Related papers  (2026-05-07T13:35:11Z)
• Disordered Dynamics in High Dimensions: Connections to Random Matrices and Machine Learning [52.26396748560348]
  We provide an overview of high dimensional dynamical systems driven by random matrices.<n>We focus on applications to simple models of learning and generalization in machine learning theory.
  arXiv  Detail & Related papers  (2026-01-03T00:12:32Z)
• Phase diagram and eigenvalue dynamics of stochastic gradient descent in multilayer neural networks [0.0]
  We study a neural network's phase diagram, in which each phase is characterised by distinctive dynamics of weight matrices.<n>Using a Langevin equation for gradient descent, we demonstrate that the three dynamical regimes can be classified effectively.
  arXiv  Detail & Related papers  (2025-09-01T10:38:31Z)
• An Analytical Characterization of Sloppiness in Neural Networks: Insights from Linear Models [18.99511760351873]
  Recent experiments have shown that training trajectories of multiple deep neural networks evolve on a remarkably low-dimensional "hyper-ribbon-like" manifold.<n>Inspired by the similarities in the training trajectories of deep networks and linear networks, we analytically characterize this phenomenon for the latter.<n>We show that the geometry of this low-dimensional manifold is controlled by (i) the decay rate of the eigenvalues of the input correlation matrix of the training data, (ii) the relative scale of the ground-truth output to the weights at the beginning of training, and (iii) the number of steps of gradient descent.
  arXiv  Detail & Related papers  (2025-05-13T19:20:19Z)
• Approaching Deep Learning through the Spectral Dynamics of Weights [41.948042468042374]
  spectral dynamics of weights -- the behavior of singular values and vectors during optimization -- to clarify and unify several phenomena in deep learning. We identify a consistent bias in optimization across various experiments, from small-scale grokking'' to large-scale tasks like image classification with ConvNets, image generation with UNets, speech recognition with LSTMs, and language modeling with Transformers.
  arXiv  Detail & Related papers  (2024-08-21T17:48:01Z)
• Machine learning in and out of equilibrium [58.88325379746631]
  Our study uses a Fokker-Planck approach, adapted from statistical physics, to explore these parallels. We focus in particular on the stationary state of the system in the long-time limit, which in conventional SGD is out of equilibrium. We propose a new variation of Langevin dynamics (SGLD) that harnesses without replacement minibatching.
  arXiv  Detail & Related papers  (2023-06-06T09:12:49Z)
• Capturing dynamical correlations using implicit neural representations [85.66456606776552]
  We develop an artificial intelligence framework which combines a neural network trained to mimic simulated data from a model Hamiltonian with automatic differentiation to recover unknown parameters from experimental data. In doing so, we illustrate the ability to build and train a differentiable model only once, which then can be applied in real-time to multi-dimensional scattering data.
  arXiv  Detail & Related papers  (2023-04-08T07:55:36Z)
• Spectral Evolution and Invariance in Linear-width Neural Networks [8.419660614226816]
  We investigate the spectral properties of linear-width feed-forward neural networks. We show that the spectra of weight in this high dimensional regime are invariant when trained by gradient descent for small constant learning rates. We also show that after adaptive gradient training, where a lower test error and feature learning emerge, both weight and kernel exhibit heavy tail behavior.
  arXiv  Detail & Related papers  (2022-11-11T23:00:30Z)
• Momentum Diminishes the Effect of Spectral Bias in Physics-Informed Neural Networks [72.09574528342732]
  Physics-informed neural network (PINN) algorithms have shown promising results in solving a wide range of problems involving partial differential equations (PDEs) They often fail to converge to desirable solutions when the target function contains high-frequency features, due to a phenomenon known as spectral bias. In the present work, we exploit neural tangent kernels (NTKs) to investigate the training dynamics of PINNs evolving under gradient descent with momentum (SGDM)
  arXiv  Detail & Related papers  (2022-06-29T19:03:10Z)
• Neural Dynamic Mode Decomposition for End-to-End Modeling of Nonlinear Dynamics [49.41640137945938]
  We propose a neural dynamic mode decomposition for estimating a lift function based on neural networks. With our proposed method, the forecast error is backpropagated through the neural networks and the spectral decomposition. Our experiments demonstrate the effectiveness of our proposed method in terms of eigenvalue estimation and forecast performance.
  arXiv  Detail & Related papers  (2020-12-11T08:34:26Z)
• Kernel and Rich Regimes in Overparametrized Models [69.40899443842443]
  We show that gradient descent on overparametrized multilayer networks can induce rich implicit biases that are not RKHS norms. We also demonstrate this transition empirically for more complex matrix factorization models and multilayer non-linear networks.
  arXiv  Detail & Related papers  (2020-02-20T15:43:02Z)

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.