{"id":"arxiv/2610.02923","url":"https://arxiv.org/abs/2610.02923","tags":["math.CO"],"score":0,"title":"Local and Global Spectral Bounds for Hermitian $A_α$-Matrices","domain":"math.CO","source":"arxiv","authors":["Ravinder Kumar","Amisha Shekhawat"],"reviews":0,"subject":"math.CO","summary":"We establish local and global spectral bounds for Hermitian $A_α$-matrices of mixed graphs. Using the first three spectral moments, we obtain an upper bound for the largest eigenvalue as the largest real zero of an explicit cubic polynomial. Vertexwise estimates for the extreme eigenvalues yield new lower bounds for the spectral spread, including a bound that strictly improves an existing degree-based estimate for ordinary graphs. A local two-dimensional compression produces a spread bound involving neighbour-degree data and gain-weighted triangles. This bound is exact for every mixed…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-10-02","reactions":{}}