Title:Entropic Regularization in the Deep Linear Network
Abstract:We study regularization for the deep linear network (DLN) using the entropy formula introduced in arXiv:2509.09088. The equilibria and gradient flow of the free energy on the Riemannian manifold of end-to-end maps of the DLN are characterized for energies that depend symmetrically on the singular values of the end-to-end matrix. The only equilibria are minimizers and the set of minimizers is an orbit of the orthogonal group. In contrast with random matrix theory there is no singular value repulsion. The corresponding gradient flow reduces to a one-dimensional ordinary d…
Title:Entropic Regularization in the Deep Linear Network
Abstract:We study regularization for the deep linear network (DLN) using the entropy formula introduced in arXiv:2509.09088. The equilibria and gradient flow of the free energy on the Riemannian manifold of end-to-end maps of the DLN are characterized for energies that depend symmetrically on the singular values of the end-to-end matrix. The only equilibria are minimizers and the set of minimizers is an orbit of the orthogonal group. In contrast with random matrix theory there is no singular value repulsion. The corresponding gradient flow reduces to a one-dimensional ordinary differential equation whose solution gives explicit relaxation rates toward the minimizers. We also study the concavity of the entropy in the chamber of singular values. The entropy is shown to be strictly concave in the Euclidean geometry on the chamber but not in the Riemannian geometry defined by the DLN metric.
| Subjects: | Neural and Evolutionary Computing (cs.NE); Dynamical Systems (math.DS); Probability (math.PR) |
| MSC classes: | 37N40, 53C20, 15A18, 68T07 |
| Cite as: | arXiv:2512.06137 [cs.NE] |
| (or arXiv:2512.06137v1 [cs.NE] for this version) | |
| https://doi.org/10.48550/arXiv.2512.06137 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Govind Menon [view email] [v1] Fri, 5 Dec 2025 20:36:13 UTC (558 KB)