{"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","summary":"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…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-01-22","reactions":{}}