PINNing the Pion: Conformal Deep Learning for Fπ(s) and the Muon g-2 Hadronic Contribution

License: See LICENSE file · Updated 2026-09-30
My rating

This paper targets high-energy and hadronic physics by applying a Physics-Informed Neural Network to the pion electromagnetic form factor and its connection to the hadronic contribution to the muon g-2. The approach works in a conformal z-plane and builds Fπ(s) from first principles across spacelike and timelike domains while incorporating charge normalization, Schwarz reflection, analyticity, and dispersion constraints.

Tags: