The Map Behind the Flow: Finite-Step Gradient Descent as a Dynamical System

• URL: http://arxiv.org/abs/2607.04993v1
• Date: Mon, 06 Jul 2026 12:28:24 GMT
• Title: The Map Behind the Flow: Finite-Step Gradient Descent as a Dynamical System
• Abstract summary: Edge-of-stability behavior, sharpness oscillations, catapult phases, balancing, and movement toward flatter representations are effects of the training map itself.<n>This paper studies fixed-step gradient descent as a discrete dynamical system in a hierarchy of exactly solvable models.
• Score: 42.117840303613164
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Many phenomena of deep learning are dynamical: they concern not only which minima exist, but how gradient descent reaches, avoids, or selects among them. Edge-of-stability behavior, sharpness oscillations, catapult phases, balancing, and movement toward flatter representations are effects of the training map itself, and are poorly captured by the small-step gradient-flow limit. This paper studies fixed-step gradient descent as a discrete dynamical system in a hierarchy of exactly solvable models retaining basic structures of deep learning: depth, factorization, width, data coupling, activation, and stochasticity. The starting point is the balanced scalar reduction of a deep linear chain, giving a quartic loss and a cubic gradient map whose post-edge behavior is explicit. Under the natural large-depth scaling, this dynamics converges to a universal Ricker-type map. The edge of stability is therefore not a breakdown of optimization, but the first bifurcation of the training map. Embedding the scalar dynamics back into factored models turns these regimes into learning phenomena. Finite steps break conservation laws of gradient flow and contract factorization imbalance; residual oscillations move parameters toward flatter, more balanced representations. Wider linear networks produce a ladder of spectral edges, so the optimal learning rate can lie beyond the first edge. Data coupling, nonlinear activations, and stochastic targets preserve the same organizing principle: finite-step oscillations drive alignment, balancing, and representation selection. Thus the learning rate is not merely a numerical stability parameter. It is a structural parameter of the training dynamics, determining its attractors and shaping the representations gradient descent selects.

Related papers

• Learning Orthogonal Multi-Index Models Beyond Small Initialization: Incremental Learning, Competitive Dynamics and Symmetry [52.75288960194919]
  Recent work has identified incremental learning in shallow networks trained on single-index and multi-index models.<n>We introduce a symmetry-based finite-width approximation via symmetrized networks, rather than comparing directly with an infinite-width limit.
  arXiv  Detail & Related papers  (2026-09-09T22:35:31Z)
• Edge Flow: A Tractable and Predictive Continuous-Time Model for Gradient Descent at the Edge of Stability [11.598128347617957]
  We propose a tractable, faithful, and predictive model of gradient descent dynamics at the edge of stability (EoS)<n>We show that Edge Flow tracks the dynamics of gradient descent at least as faithfully as previously proposed continuous-time EoS models.
  arXiv  Detail & Related papers  (2026-06-16T15:45:47Z)
• When Both Layers Learn: Training Dynamics of Representing Linear Models via ReLU Networks [22.81761236732655]
  We study the gradient descent dynamics for jointly training both layers of a one-hidden-layer ReLU network to fit a linear target function.<n>Our analysis tracks the trajectory through three phases: an alignment phase in which hidden weights progressively align with the planted direction while the output weights maintain the correct sign pattern.<n>We establish novel uniform concentration results that hold along the entire trajectory, and are essential for obtaining order-wise optimal sample complexity.
  arXiv  Detail & Related papers  (2026-06-03T05:44:30Z)
• Natural gradient descent with momentum [0.0]
  We introduce a natural version of classical inertial dynamic methods like Heavy-Ball or Nesterov.<n>This work shows how it can improve the learning process when working with nonlinear model classes.
  arXiv  Detail & Related papers  (2026-04-16T22:09:39Z)
• Gradient Flow Equations for Deep Linear Neural Networks: A Survey from a Network Perspective [0.0]
  The paper surveys recent progresses in understanding the dynamics and loss landscape of the gradient flow equations associated to deep linear neural networks.<n>The loss landscape is characterized by infinitely many global minima and saddle points, both strict and nonstrict, but lacks local minima and maxima.<n>The adjacency matrix representation we use in the paper allows to highlight the existence of a quotient space structure.
  arXiv  Detail & Related papers  (2025-11-13T14:39:32Z)
• Unlearning-based Neural Interpretations [51.99182464831169]
  We show that current baselines defined using static functions are biased, fragile and manipulable.<n>We propose UNI to compute an (un)learnable, debiased and adaptive baseline by perturbing the input towards an unlearning direction of steepest ascent.
  arXiv  Detail & Related papers  (2024-10-10T16:02:39Z)
• On the Convergence of Gradient Descent for Large Learning Rates [55.33626480243135]
  We show that convergence is impossible when a fixed step size is used.<n>We provide a proof of this in the case of linear neural networks with a squared loss.<n>We also prove the impossibility of convergence for more general losses without requiring strong assumptions such as Lipschitz continuity for the gradient.
  arXiv  Detail & Related papers  (2024-02-20T16:01:42Z)
• 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)
• On Training Implicit Models [75.20173180996501]
  We propose a novel gradient estimate for implicit models, named phantom gradient, that forgoes the costly computation of the exact gradient. Experiments on large-scale tasks demonstrate that these lightweight phantom gradients significantly accelerate the backward passes in training implicit models by roughly 1.7 times.
  arXiv  Detail & Related papers  (2021-11-09T14:40:24Z)
• Implicit Bias in Deep Linear Classification: Initialization Scale vs Training Accuracy [71.25689267025244]
  We show how the transition is controlled by the relationship between the scale and how accurately we minimize the training loss. Our results indicate that some limit behaviors of gradient descent only kick in at ridiculous training accuracies.
  arXiv  Detail & Related papers  (2020-07-13T23:49:53Z)

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.