There is a Singularity in the Loss Landscape

• URL: http://arxiv.org/abs/2201.06964v2
• Date: Fri, 19 Jul 2024 19:49:23 GMT
• Title: There is a Singularity in the Loss Landscape
• Abstract summary: We show that as the size of the dataset increases, a point forms where the magnitude of the gradient of the loss becomes unbounded. This singularity explains a variety of phenomena recently observed in the Hessian of neural network loss functions.
• Score: 0.0
• License: http://creativecommons.org/publicdomain/zero/1.0/
• Abstract: Despite the widespread adoption of neural networks, their training dynamics remain poorly understood. We show experimentally that as the size of the dataset increases, a point forms where the magnitude of the gradient of the loss becomes unbounded. Gradient descent rapidly brings the network close to this singularity in parameter space, and further training takes place near it. This singularity explains a variety of phenomena recently observed in the Hessian of neural network loss functions, such as training on the edge of stability and the concentration of the gradient in a top subspace. Once the network approaches the singularity, the top subspace contributes little to learning, even though it constitutes the majority of the gradient.

Related papers

• Phase Diagram of Dropout for Two-Layer Neural Networks in the Mean-Field Regime [20.21806452403366]
  Dropout is a training technique for neural networks that consists of randomly deactivating units at each step of their gradient-based training.<n>We study the large-widths of gradient descent with dropout on two-layer neural networks with the mean-field scale.
  arXiv  Detail & Related papers  (2025-10-08T21:09:32Z)
• The Butterfly Effect: Neural Network Training Trajectories Are Highly Sensitive to Initial Conditions [51.68215326304272]
  We show that even small perturbations reliably cause otherwise identical training trajectories to diverge-an effect that diminishes rapidly over training time.<n>Our findings provide insights into neural network training stability, with practical implications for fine-tuning, model merging, and diversity of model ensembles.
  arXiv  Detail & Related papers  (2025-06-16T08:35:16Z)
• New Evidence of the Two-Phase Learning Dynamics of Neural Networks [59.55028392232715]
  We introduce an interval-wise perspective that compares network states across a time window.<n>We show that the response of the network to a perturbation exhibits a transition from chaotic to stable.<n>We also find that after this transition point the model's functional trajectory is confined to a narrow cone-shaped subset.
  arXiv  Detail & Related papers  (2025-05-20T04:03:52Z)
• Does SGD really happen in tiny subspaces? [18.283839252425803]
  Recent studies have revealed that, along the training trajectory, the gradient approximately aligns with a low-rank top eigenspace of the training loss Hessian. This paper explores whether neural networks can be trained within the dominant subspace, which could lead to more efficient training methods.
  arXiv  Detail & Related papers  (2024-05-25T01:44:35Z)
• Early Directional Convergence in Deep Homogeneous Neural Networks for Small Initializations [2.310288676109785]
  This paper studies the gradient flow dynamics that arise when training deep homogeneous neural networks. The weights of the neural network remain small in norm and approximately converge in direction along the Karush-Kuhn-Tucker points.
  arXiv  Detail & Related papers  (2024-03-12T23:17:32Z)
• Loss Landscape of Shallow ReLU-like Neural Networks: Stationary Points, Saddle Escape, and Network Embedding [1.4513150969598634]
  We study the loss landscape of one-hidden-layer neural networks with ReLU-like activation functions trained with the empirical squared loss. We identify the stationary points of such networks, which significantly slow down loss decrease during training.
  arXiv  Detail & Related papers  (2024-02-08T12:30:29Z)
• Implicit Bias of Gradient Descent for Two-layer ReLU and Leaky ReLU Networks on Nearly-orthogonal Data [66.1211659120882]
  The implicit bias towards solutions with favorable properties is believed to be a key reason why neural networks trained by gradient-based optimization can generalize well. While the implicit bias of gradient flow has been widely studied for homogeneous neural networks (including ReLU and leaky ReLU networks), the implicit bias of gradient descent is currently only understood for smooth neural networks.
  arXiv  Detail & Related papers  (2023-10-29T08:47:48Z)
• Smooth Exact Gradient Descent Learning in Spiking Neural Networks [0.0]
  We demonstrate exact gradient descent based on continuously changing spiking dynamics. These are generated by neuron models whose spikes vanish and appear at the end of a trial, where it cannot influence subsequent dynamics.
  arXiv  Detail & Related papers  (2023-09-25T20:51:00Z)
• Implicit Bias in Leaky ReLU Networks Trained on High-Dimensional Data [63.34506218832164]
  In this work, we investigate the implicit bias of gradient flow and gradient descent in two-layer fully-connected neural networks with ReLU activations. For gradient flow, we leverage recent work on the implicit bias for homogeneous neural networks to show that leakyally, gradient flow produces a neural network with rank at most two. For gradient descent, provided the random variance is small enough, we show that a single step of gradient descent suffices to drastically reduce the rank of the network, and that the rank remains small throughout training.
  arXiv  Detail & Related papers  (2022-10-13T15:09:54Z)
• Gradient flow dynamics of shallow ReLU networks for square loss and orthogonal inputs [19.401271427657395]
  The training of neural networks by gradient descent methods is a cornerstone of the deep learning revolution. This article presents the gradient flow dynamics of one neural network for the mean squared error at small initialisation.
  arXiv  Detail & Related papers  (2022-06-02T09:01:25Z)
• Gradient Descent on Neural Networks Typically Occurs at the Edge of Stability [94.4070247697549]
  Full-batch gradient descent on neural network training objectives operates in a regime we call the Edge of Stability. In this regime, the maximum eigenvalue of the training loss Hessian hovers just above the numerical value $2 / text(step size)$, and the training loss behaves non-monotonically over short timescales, yet consistently decreases over long timescales.
  arXiv  Detail & Related papers  (2021-02-26T22:08:19Z)
• When does gradient descent with logistic loss find interpolating two-layer networks? [51.1848572349154]
  We show that gradient descent drives the training loss to zero if the initial loss is small enough. When the data satisfies certain cluster and separation conditions and the network is wide enough, we show that one step of gradient descent reduces the loss sufficiently that the first result applies.
  arXiv  Detail & Related papers  (2020-12-04T05:16:51Z)
• Gradient Starvation: A Learning Proclivity in Neural Networks [97.02382916372594]
  Gradient Starvation arises when cross-entropy loss is minimized by capturing only a subset of features relevant for the task. This work provides a theoretical explanation for the emergence of such feature imbalance in neural networks.
  arXiv  Detail & Related papers  (2020-11-18T18:52:08Z)
• Plateau Phenomenon in Gradient Descent Training of ReLU networks: Explanation, Quantification and Avoidance [0.0]
  In general, neural networks are trained by gradient type optimization methods. The loss function decreases rapidly at the beginning of training but then, after a relatively small number of steps, significantly slow down. The present work aims to identify and quantify the root causes of plateau phenomenon.
  arXiv  Detail & Related papers  (2020-07-14T17:33:26Z)
• The Break-Even Point on Optimization Trajectories of Deep Neural Networks [64.7563588124004]
  We argue for the existence of the "break-even" point on this trajectory. We show that using a large learning rate in the initial phase of training reduces the variance of the gradient. We also show that using a low learning rate results in bad conditioning of the loss surface even for a neural network with batch normalization layers.
  arXiv  Detail & Related papers  (2020-02-21T22:55:51Z)

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.