Calomplification — the power of generative calorimeter models

Bieringer, S.; Butter, Anja; Diefenbacher, Sascha; Eren, E.; Gaede, Frank; Daniel, Hundhausen,; Kasieczka, G.; Nachman, B. P.; Plehn, Tilman; Trabs, M.

doi:10.1088/1748-0221/17/09/p09028

Cited by 29 publications

(15 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As becomes apparent, the mean AUC of L2LFLOWS worsens with more showers, which is unsurprising, as with more statistics, the classifier can find more differences between the GEANT4and L2LFLOWS-generated showers. At an even larger number of showers used for classifier training, we would expect the finite size of the generator training set to become an issue, too [47][48][49]. Nevertheless, we observe that for a given number of showers, the BIB-AE showers are more separable from GEANT4 than the L2LFLOWS showers, indicating a better performance of L2LFLOWS.…”

Section: Classifier Testsmentioning

confidence: 78%

L2LFlows: Generating High-Fidelity 3D Calorimeter Images

Diefenbacher¹,

Eren²,

Gaede³

et al. 2023

Preprint

View full text Add to dashboard Cite

We explore the use of normalizing flows to emulate Monte Carlo detector simulations of photon showers in a high-granularity electromagnetic calorimeter prototype for the International Large Detector (ILD). Our proposed method -which we refer to as "Layer-to-Layer-Flows" (L2LFLOWS) -is an evolution of the CaloFlow architecture adapted to a higher-dimensional setting (30 layers of 10 × 10 voxels each). The main innovation of L2LFLOWS consists of introducing 30 separate normalizing flows, one for each layer of the calorimeter, where each flow is conditioned on the previous five layers in order to learn the layer-to-layer correlations. We compare our results to the BIB-AE, a state-of-the-art generative network trained on the same dataset and find a significantly improved fidelity of our model.

show abstract

Section: Classifier Testsmentioning

confidence: 78%

L2LFlows: Generating High-Fidelity 3D Calorimeter Images

Diefenbacher¹,

Eren²,

Gaede³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Given the interpolation properties of neural networks and the benefits of their implicit bias in the applications described in Sec. 2, we can quantify the amplification of statistics-limited training data through generative networks [65,66].…”

Section: End-to-end Ml-generatorsmentioning

confidence: 99%

“…Neural networks work much like a fit and not like an interpolation in the sense that they do not reproduce the training data faithfully and instead learn a smooth approximation [65,66]. This is where we can gain some intuition for a NN-uncertainty treatment.…”

Section: Control and Precisionmentioning

confidence: 99%

Machine learning and LHC event generation

et al. 2023

View full text Add to dashboard Cite

First-principle simulations are at the heart of the high-energy physics research program. They link the vast data output of multi-purpose detectors with fundamental theory predictions and interpretation. This review illustrates a wide range of applications of modern machine learning to event generation and simulation-based inference, including conceptional developments driven by the specific requirements of particle physics. New ideas and tools developed at the interface of particle physics and machine learning will improve the speed and precision of forward simulations, handle the complexity of collision data, and enhance inference as an inverse simulation problem.

show abstract

“…For these applications, the generative models need to learn the underlying data distributions with the help of a strong inductive bias on the model architecture. In this way, it is possible to amplify the training statistics [2,3] as well as utilize generative modeling as a powerful data augmentation technique [4].…”

Section: Introductionmentioning

confidence: 99%

EPiC-GAN: Equivariant Point Cloud Generation for Particle Jets

Buhmann¹,

Kasieczka²,

Thaler³

2023

Preprint

View full text Add to dashboard Cite

With the vast data-collecting capabilities of current and future high-energy collider experiments, there is an increasing demand for computationally efficient simulations. Generative machine learning models enable fast event generation, yet so far these approaches are largely constrained to fixed data structures and rigid detector geometries. In this paper, we introduce EPiC-GAN -equivariant point cloud generative adversarial network -which can produce point clouds of variable multiplicity. This flexible framework is based on deep sets and is well suited for simulating sprays of particles called jets. The generator and discriminator utilize multiple EPiC layers with an interpretable global latent vector. Crucially, the EPiC layers do not rely on pairwise information sharing between particles, which leads to a significant speedup over graph-and transformer-based approaches with more complex relation diagrams. We demonstrate that EPiC-GAN scales well to large particle multiplicities and achieves high generation fidelity on benchmark jet generation tasks.

show abstract

Calomplification — the power of generative calorimeter models

Cited by 29 publications

References 45 publications

L2LFlows: Generating High-Fidelity 3D Calorimeter Images

L2LFlows: Generating High-Fidelity 3D Calorimeter Images

Machine learning and LHC event generation

EPiC-GAN: Equivariant Point Cloud Generation for Particle Jets

Contact Info

Product

Resources

About