Depth, Not Data: An Analysis of Hessian Spectral Bifurcation

• URL: http://arxiv.org/abs/2602.00545v1
• Date: Sat, 31 Jan 2026 06:17:00 GMT
• Title: Depth, Not Data: An Analysis of Hessian Spectral Bifurcation
• Authors: Shenyang Deng, Boyao Liao, Zhuoli Ouyang, Tianyu Pang, Yaoqing Yang,
• Abstract summary: The eigenvalue distribution of the Hessian matrix plays a crucial role in understanding the landscape of deep neural networks.<n>We demonstrate that such spectral Bifurcation can arise purely from the network architecture, independent of data imbalance.<n>Our results suggest that both model architecture and data characteristics should be considered when designing optimization algorithms for deep networks.
• Score: 29.426396222985563
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The eigenvalue distribution of the Hessian matrix plays a crucial role in understanding the optimization landscape of deep neural networks. Prior work has attributed the well-documented ``bulk-and-spike'' spectral structure, where a few dominant eigenvalues are separated from a bulk of smaller ones, to the imbalance in the data covariance matrix. In this work, we challenge this view by demonstrating that such spectral Bifurcation can arise purely from the network architecture, independent of data imbalance. Specifically, we analyze a deep linear network setup and prove that, even when the data covariance is perfectly balanced, the Hessian still exhibits a Bifurcation eigenvalue structure: a dominant cluster and a bulk cluster. Crucially, we establish that the ratio between dominant and bulk eigenvalues scales linearly with the network depth. This reveals that the spectral gap is strongly affected by the network architecture rather than solely by data distribution. Our results suggest that both model architecture and data characteristics should be considered when designing optimization algorithms for deep networks.

Related papers

• Hodge-Aware Contrastive Learning [101.56637264703058]
  Simplicial complexes prove effective in modeling data with multiway dependencies. We develop a contrastive self-supervised learning approach for processing simplicial data.
  arXiv  Detail & Related papers  (2023-09-14T00:40:07Z)
• A theory of data variability in Neural Network Bayesian inference [0.70224924046445]
  We provide a field-theoretic formalism which covers the generalization properties of infinitely wide networks. We derive the generalization properties from the statistical properties of the input. We show that data variability leads to a non-Gaussian action reminiscent of a ($varphi3+varphi4$)-theory.
  arXiv  Detail & Related papers  (2023-07-31T14:11:32Z)
• Bayesian Interpolation with Deep Linear Networks [92.1721532941863]
  Characterizing how neural network depth, width, and dataset size jointly impact model quality is a central problem in deep learning theory. We show that linear networks make provably optimal predictions at infinite depth. We also show that with data-agnostic priors, Bayesian model evidence in wide linear networks is maximized at infinite depth.
  arXiv  Detail & Related papers  (2022-12-29T20:57:46Z)
• The SVD of Convolutional Weights: A CNN Interpretability Framework [3.5783190448496343]
  We propose a framework against which interpretability methods might be applied using hypergraphs to model class separation. Rather than looking to the activations to explain the network, we use the singular vectors with the greatest corresponding singular values for each linear layer to identify those features most important to the network.
  arXiv  Detail & Related papers  (2022-08-14T18:23:02Z)
• coVariance Neural Networks [119.45320143101381]
  Graph neural networks (GNN) are an effective framework that exploit inter-relationships within graph-structured data for learning. We propose a GNN architecture, called coVariance neural network (VNN), that operates on sample covariance matrices as graphs. We show that VNN performance is indeed more stable than PCA-based statistical approaches.
  arXiv  Detail & Related papers  (2022-05-31T15:04:43Z)
• Optimization-Based Separations for Neural Networks [57.875347246373956]
  We show that gradient descent can efficiently learn ball indicator functions using a depth 2 neural network with two layers of sigmoidal activations. This is the first optimization-based separation result where the approximation benefits of the stronger architecture provably manifest in practice.
  arXiv  Detail & Related papers  (2021-12-04T18:07:47Z)
• Does the Data Induce Capacity Control in Deep Learning? [0.0]
  This paper studies how the dataset may be the cause of the anomalous generalization performance of deep networks. We show that the data correlation matrix of typical classification datasets has an eigenspectrum where, after a sharp initial drop, a large number of small eigenvalues are distributed uniformly over an exponentially large range.
  arXiv  Detail & Related papers  (2021-10-27T04:40:27Z)
• ReduNet: A White-box Deep Network from the Principle of Maximizing Rate Reduction [32.489371527159236]
  This work attempts to provide a plausible theoretical framework that aims to interpret modern deep (convolutional) networks from the principles of data compression and discriminative representation. We show that for high-dimensional multi-class data, the optimal linear discriminative representation maximizes the coding rate difference between the whole dataset and the average of all the subsets. We show that the basic iterative gradient ascent scheme for optimizing the rate reduction objective naturally leads to a multi-layer deep network, named ReduNet, that shares common characteristics of modern deep networks.
  arXiv  Detail & Related papers  (2021-05-21T16:29:57Z)
• The Heterogeneity Hypothesis: Finding Layer-Wise Differentiated Network Architectures [179.66117325866585]
  We investigate a design space that is usually overlooked, i.e. adjusting the channel configurations of predefined networks. We find that this adjustment can be achieved by shrinking widened baseline networks and leads to superior performance. Experiments are conducted on various networks and datasets for image classification, visual tracking and image restoration.
  arXiv  Detail & Related papers  (2020-06-29T17:59:26Z)
• Revealing the Structure of Deep Neural Networks via Convex Duality [70.15611146583068]
  We study regularized deep neural networks (DNNs) and introduce a convex analytic framework to characterize the structure of hidden layers. We show that a set of optimal hidden layer weights for a norm regularized training problem can be explicitly found as the extreme points of a convex set. We apply the same characterization to deep ReLU networks with whitened data and prove the same weight alignment holds.
  arXiv  Detail & Related papers  (2020-02-22T21:13:44Z)

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.