Learning S-Matrix Phases with Neural Operators

TL;DR

This paper employs Fourier Neural Operators to learn the relationship between amplitude modulus and phase in elastic scattering, bypassing traditional unitarity constraints, and introduces a novel regression-classification approach to assess phase prediction quality.

Contribution

It presents a new application of FNOs to discover unitarity relations in scattering amplitudes without explicit integral constraints, combining regression and classification tasks.

Findings

FNOs accurately predict phases of amplitudes with infinite partial waves

The classification index correlates with unitarity violations

The method delineates allowed and disallowed amplitude profiles

Abstract

We use Fourier Neural Operators (FNOs) to study the relation between the modulus and phase of amplitudes in $2\to 2$ elastic scattering at fixed energies. Unlike previous approaches, we do not employ the integral relation imposed by unitarity, but instead train FNOs to discover it from many samples of amplitudes with finite partial wave expansions. When trained only on true samples, the FNO correctly predicts (unique or ambiguous) phases of amplitudes with infinite partial wave expansions. When also trained on false samples, it can rate the quality of its prediction by producing a true/false classifying index. We observe that the value of this index is strongly correlated with the violation of the unitarity constraint for the predicted phase, and present examples where it delineates the boundary between allowed and disallowed profiles of the modulus. Our application of FNOs is unconventional: it involves a simultaneous regression-classification task and emphasizes the role of statistics in ensembles of NOs. We comment on the merits and limitations of the approach and its potential as a new methodology in Theoretical Physics.