Even Faster Simulations with Flow Matching: A Study of Zero Degree Calorimeter Responses

2025-07-24

AI Summary

High-fidelity simulation of the ALICE Zero Degree Calorimeter (ZDC) incurs prohibitive computational cost, hindering real-time analysis in high-energy physics (HEP). Method: This work introduces flow matching—a novel generative modeling paradigm—to HEP detector response modeling for the first time, constructing a low-parameter, high-fidelity surrogate model. Our approach integrates latent-space representation with an efficient training strategy, ensuring physical consistency while drastically accelerating inference. Contribution/Results: The model achieves Wasserstein distances of 1.27 and 1.30 for neutron and proton detector responses, respectively—surpassing the prior state-of-the-art (2.08). Single-sample inference time is reduced to 0.46 ms, and latent-space sampling further drops to 0.026 ms. By reconciling accuracy and speed, this work establishes a new paradigm for real-time HEP simulation.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Search and Optimization: Sampling/Simulation-based Search
• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior

Application Category

• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Web Mining and Content Analysis: Web data generation and simulation

Abstract

Recent advances in generative neural networks, particularly flow matching (FM), have enabled the generation of high-fidelity samples while significantly reducing computational costs. A promising application of these models is accelerating simulations in high-energy physics (HEP), helping research institutions meet their increasing computational demands. In this work, we leverage FM to develop surrogate models for fast simulations of zero degree calorimeters in the ALICE experiment. We present an effective training strategy that enables the training of fast generative models with an exceptionally low number of parameters. This approach achieves state-of-the-art simulation fidelity for both neutron (ZN) and proton (ZP) detectors, while offering substantial reductions in computational costs compared to existing methods. Our FM model achieves a Wasserstein distance of 1.27 for the ZN simulation with an inference time of 0.46 ms per sample, compared to the current best of 1.20 with an inference time of approximately 109 ms. The latent FM model further improves the inference speed, reducing the sampling time to 0.026 ms per sample, with a minimal trade-off in accuracy. Similarly, our approach achieves a Wasserstein distance of 1.30 for the ZP simulation, outperforming the current best of 2.08. The source code is available at https://github.com/m-wojnar/faster_zdc.

Problem

• Accelerate zero degree calorimeter simulations using flow matching
• Reduce computational costs in high-energy physics experiments
• Improve simulation fidelity for neutron and proton detectors

Innovation

• Flow matching for fast calorimeter simulations
• Low-parameter generative models training strategy
• Latent FM model boosts inference speed