{"id":"hexagon/2611.00662","pdf":"https://hexagonmath.org/pdf/2611.00662v1","url":"https://hexagonmath.org/abs/2611.00662v1","tags":["math.ST","msc:60B20","msc:15B52","msc:62H25","Gaussian mixtures","exact zero weights","random matrices","rectified linear unit","spectral outliers","weighted sample covariance matrices"],"score":0,"title":"Spectral outliers of self-coupled covariance matrices with exact zero weights","domain":"math.PR","origin":"ai-generated","source":"hexagon","authors":["Samuil Petkov"],"reviews":0,"subject":"math.PR","summary":"We study the right spectral outliers of $M_d=n^{-1}A_dDA_d^{\\mathsf T}$, where the columns of $A_d$ are drawn from a Gaussian mixture with a fixed-dimensional signal subspace and each diagonal weight $D_{\\ell\\ell}=h_{b_\\ell}(g_\\ell)$ is a function of the class label and of the low-dimensional Gaussian coordinates of the same sample. The effective outlier theory of Ben Arous, Gheissari, Huang and Jagannath for such self-coupled matrices assumes that the weights do not accumulate at zero, which excludes hard gates such as the derivative of a rectified linear unit. Assuming…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"ai_models":["ChatGPT","Codex","Claude Code"],"discussed":0,"published":"2026-11-19","reactions":{},"withdrawn":false,"authorship":"mixed","formalized":false,"human_understanding":"all"}