Calomplification -- The Power of Generative Calorimeter Models

Abstract

Motivated by the high computational costs of classical simulations, machine-learned generative models can be extremely useful in particle physics and elsewhere. They become especially attractive when surrogate models can efficiently learn the underlying distribution, such that a generated sample outperforms a training sample of limited size. This kind of GANplification has been observed for simple Gaussian models. We show the same effect for a physics simulation, specifically photon showers in an electromagnetic calorimeter.

17 pages, 10 figures

References in corpus (5)

• SymmetryGAN: Symmetry Discovery with Deep Learning
• A method to challenge symmetries in data with self-supervised learning
• Comparing Machine Learning and Interpolation Methods for Loop-Level Calculations
• Reduced Precision Strategies for Deep Learning: A High Energy Physics Generative Adversarial Network Use Case
• Physics Validation of Novel Convolutional 2D Architectures for Speeding Up High Energy Physics Simulations

Cited by in corpus (10)

• Machine Learning and LHC Event Generation
• Score-based Generative Models for Calorimeter Shower Simulation
• Event Generators for High-Energy Physics Experiments
• Evaluating generative models in high energy physics
• EPiC-GAN: Equivariant Point Cloud Generation for Particle Jets
• Towards a Deep Learning Model for Hadronization
• Unsupervised Quantum Circuit Learning in High Energy Physics
• Accurate Surrogate Amplitudes with Calibrated Uncertainties
• Amplitude Uncertainties Everywhere All at Once
• Extrapolating Jet Radiation with Autoregressive Transformers