Physics-Informed Neural Networks for Solving Two-Flavor Neutrino Oscillations in Vacuum and Matter Environments for Atmospheric and Reactor Neutrinos Preprint: APS/123-QED

T. Srinivasan, Kalyani Desikan · 2026

Read Physics-Informed Neural Networks for Solving Two-Flavor Neutrino Oscillations in Vacuum and Matter Environments for Atmospheric and Reactor Neutrinos Preprint: APS/123-QED by T. Srinivasan, Kalyani Desikan (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-22862.txt.

Description

Srinivasan and Desikan test physics-informed neural networks, or PINNs, as solvers for two-flavor neutrino oscillation equations. The cases cover vacuum and constant-density matter, for reactor and atmospheric neutrinos. Ordinary solvers already handle these equations to machine precision cheaply, so the paper is a proof of concept. It benchmarks PINNs against NuFast and Runge-Kutta solutions. Network tuning yields high-precision agreement, setting a baseline for harder cases such as varying density profiles.

Catalog record

CREATORT. Srinivasan, Kalyani Desikan
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE27514
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.22862

Browse the full library · About the Tower of Babel