{"id":"arxiv/2610.09919","url":"https://arxiv.org/abs/2610.09919","tags":["cs.LG","math.OC","stat.ML"],"score":0,"title":"Eigenvalues of the Hessian in Deep Learning: The Origin of Symmetry and Its Breaking","domain":"cs.LG","source":"arxiv","authors":["Yossi Arjevani"],"reviews":0,"subject":"cs.LG","summary":"Hessian spectra at trained models in deep learning exhibit a persistent pattern: eigenvalues organize into distinct clusters, including a large bulk near zero and a few isolated outliers. This paper shows that a natural account of these spectral phenomena emerges when the original setting is understood as a departure from a nearby, otherwise hidden, highly symmetric reference. Modifications, including changes to the architecture, data distribution, or parameter metric, expose a nearby reference configuration whose Hessian exhibits rich invariances-ones not accounted for by weight symmetries.…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-10-07","reactions":{}}