Invertible Surrogate Models: Joint surrogate modelling and reconstruction of Laser-Wakefield Acceleration by invertible neural networks

TL;DR

This paper introduces invertible neural networks as surrogate models for simulating laser plasma accelerators, enabling both accurate approximation and reconstruction of experimental diagnostics.

Contribution

It presents a novel invertible surrogate modeling approach for laser wakefield acceleration, combining simulation approximation with diagnostic reconstruction.

Findings

High accuracy in simulating LWFA physics

Effective reconstruction of experimental diagnostics

Potential for bidirectional modeling in plasma physics

Abstract

Invertible neural networks are a recent technique in machine learning promising neural network architectures that can be run in forward and reverse mode. In this paper, we will be introducing invertible surrogate models that approximate complex forward simulation of the physics involved in laser plasma accelerators: iLWFA. The bijective design of the surrogate model also provides all means for reconstruction of experimentally acquired diagnostics. The quality of our invertible laser wakefield acceleration network will be verified on a large set of numerical LWFA simulations.