Saved in:
Bibliographic Details
Main Authors: Menon, Govind, Yu, Tianmin
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2509.09088
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914583861526528
author Menon, Govind
Yu, Tianmin
author_facet Menon, Govind
Yu, Tianmin
contents We study the Riemannian geometry of the Deep Linear Network (DLN) as a foundation for a thermodynamic description of the learning process. The main tools are the use of group actions to analyze overparametrization and the use of Riemannian submersion from the space of parameters to the space of observables. The foliation of the balanced manifold in the parameter space by group orbits is used to define and compute a Boltzmann entropy. We also show that the Riemannian geometry on the space of observables defined in [2] is obtained by Riemannian submersion of the balanced manifold. The main technical step is an explicit construction of an orthonormal basis for the tangent space of the balanced manifold using the theory of Jacobi matrices.
format Preprint
id arxiv_https___arxiv_org_abs_2509_09088
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle An entropy formula for the Deep Linear Network
Menon, Govind
Yu, Tianmin
Machine Learning
Differential Geometry
Dynamical Systems
68T07, 49J40, 94A17, 60B20
We study the Riemannian geometry of the Deep Linear Network (DLN) as a foundation for a thermodynamic description of the learning process. The main tools are the use of group actions to analyze overparametrization and the use of Riemannian submersion from the space of parameters to the space of observables. The foliation of the balanced manifold in the parameter space by group orbits is used to define and compute a Boltzmann entropy. We also show that the Riemannian geometry on the space of observables defined in [2] is obtained by Riemannian submersion of the balanced manifold. The main technical step is an explicit construction of an orthonormal basis for the tangent space of the balanced manifold using the theory of Jacobi matrices.
title An entropy formula for the Deep Linear Network
topic Machine Learning
Differential Geometry
Dynamical Systems
68T07, 49J40, 94A17, 60B20
url https://arxiv.org/abs/2509.09088