{"id":"arxiv/2610.21964","url":"https://arxiv.org/abs/2610.21964","tags":["cs.LG","cs.AI","math.SP","stat.ML"],"score":0,"title":"Why Shallow Neural Networks Learn Sinusoidal Components First: An Orthogonal Polynomial Dynamics Framework","domain":"cs.LG","source":"arxiv","authors":["Yanyang Xiao"],"reviews":0,"subject":"cs.LG","summary":"Neural networks often learn low-frequency components before high-frequency ones --- a phenomenon called spectral bias or termed F-principle. Many numerical studies have demonstrated this phenomenon, and theoretical studies have sought to explain it. However, several goals remain insufficiently addressed: (1) explaining spectral bias asymptotically, including its origin and why the preferred components are generally sinusoidal; (2) extending the explanation to general activation functions, such as sigmoid and ReLU; (3) making weak assumptions about the data distribution; (4) avoiding aliasing…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-10-15","reactions":{}}