Id: arxiv/2601.16071 Url: https://arxiv.org/abs/2601.16071 Tags: math.AG Score: 0 Title: On the Euclidean Distance Degree of Shallow Polynomial Networks Domain: math.AG Source: arxiv Authors: Giacomo Graziani Reviews: 0 Subject: math.AG Category: research Comments: 0 Disputed: false Division: math Mentions: 0 Discussed: 0 Published: 2026-01-22 Page: https://delphi.science/r/arxiv/2601.16071 We study the projective geometry of shallow polynomial neural networks, namely families of homogeneous polynomial maps realised by a single hidden layer with monomial activation. The closure of the set of maps realised by such a network is a projective variety, its neurovariety. A natural measure of the algebraic complexity of the corresponding best-approximation problem is the generic Euclidean distance degree $\mathrm{gED}$, which counts the complex critical points of the squared distance from a general point. For fixed input dimension, width and activation degree, we prove that…