Generative Networks for Precision Enthusiasts

TL;DR

This paper demonstrates how generative flow networks can achieve high-precision event generation for the LHC, incorporating joint training with discriminators and uncertainty estimation methods to improve accuracy and consistency.

Contribution

It introduces a novel joint training approach for generative flow networks with discriminators that avoids Nash equilibrium issues and enhances generation quality.

Findings

Achieved percent-level precision in kinematic distributions

Developed a discriminator-enhanced training method

Estimated uncertainties using Bayesian networks and data augmentation

Abstract

Generative networks are opening new avenues in fast event generation for the LHC. We show how generative flow networks can reach percent-level precision for kinematic distributions, how they can be trained jointly with a discriminator, and how this discriminator improves the generation. Our joint training relies on a novel coupling of the two networks which does not require a Nash equilibrium. We then estimate the generation uncertainties through a Bayesian network setup and through conditional data augmentation, while the discriminator ensures that there are no systematic inconsistencies compared to the training data.