{"id":"arxiv/2604.20061","url":"https://arxiv.org/abs/2604.20061","tags":["math-ph","nlin.CD"],"score":0,"title":"Predictivity and Utility of Neural Surrogates of Multiscale PDEs","domain":"math-ph","source":"arxiv","authors":["Karthik Duraisamy"],"reviews":0,"subject":"math-ph","summary":"Scientific machine learning is increasingly being spoken of as universal emulators for classical numerical solvers for multi-scale partial differential equations, but most apparent successes can be explained by facts that also define their limits. Many successful benchmarks live on low-dimensional solution manifolds where any competent reduced model will interpolate well. More fundamentally, neural surrogates systematically under-resolve high-frequency content due to spectral bias, and coarse-graining compounds this problem through irreversible information loss. In many multi-scale problems,…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-04-21","reactions":{}}