Machine learning-based event generator for electron-proton scattering

We present a new machine learning-based Monte Carlo event generator using generative adversarial networks (GANs) that can be trained with calibrated detector simulations to construct vertex-level free of theoretical assumptions about femtometer scale physics. Our framework includes GAN-based folding as fast-surrogate model mimics simulators. The is tested and validated on simulated inclusive deep-inelastic scattering data along existing parametrizations for simulation, uncertainty quantification based statistical bootstrapping technique. results provide the first time realistic proof concept mitigate theory bias in inferring distributions needed reconstruct physical observables.