English

Physics informed operator learning of parameter dependent spectra

General Relativity and Quantum Cosmology 2026-04-28 v1 Instrumentation and Methods for Astrophysics Computational Physics

Abstract

Spectral problems governed by differential operators underpin a wide range of physical systems, yet remain computationally challenging because their spectra depend sensitively on continuous parameters and often demand repeated evaluations across parameter space. Here we present DeepOPiraKAN\texttt{DeepOPiraKAN}, an open source physics informed neural network architecture for spectral analysis. By combining operator learning with enhanced optimization stability, it captures the underlying parameter-to-spectrum mapping in a single model, avoiding repeated spectral solutions at isolated points in parameter space. As a representative and stringent benchmark, we apply this framework to the computation of quasinormal modes of Kerr black holes. A single trained network accurately resolves modes with (,m){(2,0),(2,1)}(\ell,m)\in \{(2,0),(2,1)\} and overtones up to n=7n=7 across the full spin range, achieving relative errors of O(106)\mathcal{O}(10^{-6}) for the fundamental mode and gradually increasing to O(104)\mathcal{O}(10^{-4}) for higher overtones, benchmarked against the Leaver's method. This level of accuracy is already significant for black hole spectroscopy and practical ringdown modelling for current and future observatories. More broadly, these results highlight the potential of DeepOPiraKAN\texttt{DeepOPiraKAN} as a general and scalable framework for parameter dependent spectral problems across complex physical systems.

Keywords

Cite

@article{arxiv.2604.23625,
  title  = {Physics informed operator learning of parameter dependent spectra},
  author = {Haohao Gu and Sensen He and Hanlin Song and Bo Liang and Zhenwei Lyu and Xiaoguang Hu and Minghui Du and Peng Xu and Bo-Qiang Ma},
  journal= {arXiv preprint arXiv:2604.23625},
  year   = {2026}
}
R2 v1 2026-07-01T12:35:39.088Z