A First Full Physics Benchmark for Highly Granular Calorimeter Surrogates

2025-11-21

AI Summary

High-granularity calorimeter (HGCAL) simulation is computationally expensive, hindering its scalability for future collider physics analyses. Method: This work introduces the first full-stack generative surrogate modeling evaluation framework tailored to realistic physics scenarios. Leveraging the DDML library, we establish a standardized interface enabling end-to-end integration of generative models with DD4hep-based detector geometry and response, thereby decoupling methodology development from physics performance assessment. Contribution/Results: We systematically benchmark both regular-grid and point-cloud representations across multi-level physics tasks—including single-particle response, diphoton separation, and τ-lepton hadronic decays—using an ideal sampler as reference. Experiments show that point-cloud models achieve GeV-scale energy resolution (<2%) while accelerating inference by 3–5× over grid-based counterparts. This breakthrough in the accuracy–efficiency trade-off establishes a new paradigm for real-time HGCAL simulation and downstream physics analysis.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Search and Optimization: Sampling/Simulation-based Search
• Natural Language Processing: Generation

Application Category

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

Abstract

The physics programs of current and future collider experiments necessitate the development of surrogate simulators for calorimeter showers. While much progress has been made in the development of generative models for this task, they have typically been evaluated in simplified scenarios and for single particles. This is particularly true for the challenging task of highly granular calorimeter simulation. For the first time, this work studies the use of highly granular generative calorimeter surrogates in a realistic simulation application. We introduce DDML, a generic library which enables the combination of generative calorimeter surrogates with realistic detectors implemented using the DD4hep toolkit. We compare two different generative models - one operating on a regular grid representation, and the other using a less common point cloud approach. In order to disentangle methodological details from model performance, we provide comparisons to idealized simulators which directly sample representations of different resolutions from the full simulation ground-truth. We then systematically evaluate model performance on post-reconstruction benchmarks for electromagnetic shower simulation. Beginning with a typical single particle study, we introduce a first multi-particle benchmark based on di-photon separations, before studying a first full-physics benchmark based on hadronic decays of the tau lepton. Our results indicate that models operating on a point cloud can achieve a favorable balance between speed and accuracy for highly granular calorimeter simulation compared to those which operate on a regular grid representation.

Problem

• Developing surrogate simulators for highly granular calorimeter showers in collider experiments
• Evaluating generative models in realistic scenarios beyond single particle simulations
• Comparing point cloud and grid-based approaches for speed and accuracy balance

Innovation

• Combining generative models with realistic detector simulations
• Comparing grid-based and point cloud generative approaches
• Evaluating models on full-physics multi-particle benchmark scenarios