CaloTrilogy: Toward a Breakthrough in One-Step, End-to-End, Physics-Guided Shower Generation for Modern Calorimeters

2026-06-02

AI Summary

High-fidelity calorimeter shower simulation is computationally expensive, and existing generative models struggle to balance physical fidelity with inference efficiency. This work proposes a unified generative framework that integrates single-step flow matching sampling, a learnable shower-space prior, and a physics-guided loss function, embedding physical constraints directly into training. The method achieves, for the first time, end-to-end generation of high-dimensional, physically consistent showers in only one to a few inference steps—without requiring additional networks or computational overhead. Evaluated on multiple public high-granularity calorimeter datasets, the approach matches the generation quality of state-of-the-art diffusion and flow-based models while significantly accelerating inference and preserving inter-layer physical structure consistency.

Abstract

High-precision calorimeter simulation at current and future colliders imposes rapidly growing computational demands, motivating the development of machine-learning surrogates for traditional Monte Carlo tools such as Geant4. Flow matching and diffusion-based generative models have become leading approaches for high-dimensional fast simulation because of their sample quality, but typically require ${\cal O}(100)$ function evaluations at inference and often rely on auxiliary networks to constrain global observables, compromising streamlined end-to-end generation. We introduce a unified framework that improves the balance between speed, shower quality, and physics fidelity. The method combines: (i) an average velocity field integrator that enables sampling in one or a few evaluations; (ii) a learned generative prior in shower space, constructed from data rather than random noise; and (iii) physics-guided loss terms that impose inductive biases on key observables during training. These elements are training time regularizers, preserving end-to-end inference with no additional cost. With only one or a few evaluation steps, the model achieves shower quality competitive with state-of-the-art flow and diffusion approaches, tested on several public high granularity calorimeter datasets. The results demonstrate inter-layer shower structure consistent with the underlying physics, providing a strong candidate for future fast simulation workflows.

Problem

• calorimeter simulation
• fast simulation
• generative models
• physics-guided generation
• computational efficiency

Innovation

• flow matching
• physics-guided generation
• fast calorimeter simulation
• generative prior
• one-step sampling