Transferable Fast Calorimeter Shower Generation via Multi-Geometry Pre-training
physics.ins-det, hep-ex, hep-ph
Submitted: 2026-08-18
Updated: 2026-10-29
Comments: 50 pages, 29 figures, 12 tables. Published in JINST 21 (2026) P10041
Journal ref: JINST 21 (2026) P10041
DOI: 10.1088/1748-0221/21/10/P10041
License: http://creativecommons.org/licenses/by/4.0/
The gist: Detailed Geant4 simulation of calorimeter showers dominates the computing budget of high-energy physics experiments.
Terminology
Abstract
Detailed Geant4 simulation of calorimeter showers dominates the computing budget of high-energy physics experiments. Deep generative surrogates reduce this cost, but they have remained tied to the detector they were trained on, so each new geometry needs a large in-domain dataset. We study whether a single point cloud shower generator can be pre-trained on multiple detectors and transferred to unseen calorimeters. The pre-training geometries come from synthetic geometric variation rather than real-detector data. We introduce SimpleBox, a family of 10 4 box calorimeters spanning the plane of sampling fraction and longitudinal segmentation, and benchmark it against pre-training on realistic detectors. On a calorimeter unseen in pre-training, with 10 cubed target showers for fine-tuning, the two priors reduce the aggregated sliced Wasserstein distance to Geant4 by factors of 5.2 (synthetic) and 8.0 (realistic) relative to training from scratch. At larger target sizes the synthetic prior performs better than the realistic one. Geometric diversity alone is therefore a practical way to pre-train a transferable shower generator.
Sources
- A Roadmap for HEP Software and Computing R&D for the 2020s
- AtlFast3: the next generation of fast simulation in ATLAS
- Accelerating Science with Generative Adversarial Networks: An Application to 3D Particle Showers in Multi-Layer Calorimeters
- CaloGAN: Simulating 3D High Energy Particle Showers in Multi-Layer Electromagnetic Calorimeters with Generative Adversarial Networks
- Controlling Physical Attributes in GAN-Accelerated Simulation of Electromagnetic Calorimeters
- Generating and refining particle detector simulations using the Wasserstein distance in adversarial networks
- Precise simulation of electromagnetic calorimeter showers using a Wasserstein Generative Adversarial Network
- Fast and accurate simulation of particle detectors using generative adversarial networks
- Calorimetry with Deep Learning: Particle Simulation and Reconstruction for Collider Physics
- Getting High: High Fidelity Simulation of High Granularity Calorimeters with High Speed
- GANplifying Event Samples
- CaloShowerGAN, a Generative Adversarial Networks model for fast calorimeter shower simulation
- CALPAGAN: Calorimetry for Particles using GANs
- Hadrons, Better, Faster, Stronger
- Deep generative models for fast photon shower simulation in ATLAS
- CaloMan: Fast generation of calorimeter showers with density estimation on learned manifolds
- CaloQVAE : Simulating high-energy particle-calorimeter interactions using hybrid quantum-classical generative models
- New Angles on Fast Calorimeter Shower Simulation
- OmniJet-${\alpha}_C$: Learning point cloud calorimeter simulations using generative transformers
- CaloFlow: Fast and Accurate Generation of Calorimeter Showers with Normalizing Flows
Related papers
- Cryogenic Mechanical Loss of GaAs/AlGaAs Crystalline Coatings
- The physical design of the superfluid helium-4 UCN source at CSNS
- A new diffuse reflector filament for additive manufacturing of 3D printing finely-segmented plastic scintillator
- Characterization of CRYO ASIC for charge readout in the nEXO experiment
- Reconstructing High-Fidelity Light Yield Maps for Surface LArTPCs Using Crossing Cosmic Muons
- OASIS-ERIS: Open Acquisition System for IEPE Sensors - Electrically Refined and Inter-device Synchronization