Saved in:
Bibliographic Details
Main Author: Duraisamy, Karthik
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2604.20061
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911613960847360
author Duraisamy, Karthik
author_facet Duraisamy, Karthik
contents 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, no architecture or training procedure can fully recover what the coarse representation discards. Two simple examples are used to characterize spectral bias, coarse-graining and error accumulation. We discuss why medium-range weather prediction on reanalysis data sits in a favorable sweet spot and why this will not generalize to genuinely chaotic multi-scale scenarios. We identify domains where neural surrogates offer genuine value, propose further research on neural-classical hybrids, and call for better reporting standards.
format Preprint
id arxiv_https___arxiv_org_abs_2604_20061
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Predictivity and Utility of Neural Surrogates of Multiscale PDEs
Duraisamy, Karthik
Mathematical Physics
Chaotic Dynamics
35
I.2.6
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, no architecture or training procedure can fully recover what the coarse representation discards. Two simple examples are used to characterize spectral bias, coarse-graining and error accumulation. We discuss why medium-range weather prediction on reanalysis data sits in a favorable sweet spot and why this will not generalize to genuinely chaotic multi-scale scenarios. We identify domains where neural surrogates offer genuine value, propose further research on neural-classical hybrids, and call for better reporting standards.
title Predictivity and Utility of Neural Surrogates of Multiscale PDEs
topic Mathematical Physics
Chaotic Dynamics
35
I.2.6
url https://arxiv.org/abs/2604.20061