Inductive Simulation of Calorimeter Showers with Normalizing Flows
This addresses the expensive simulation step in Large Hadron Collider pipelines, enabling more efficient future detector upgrades.
The paper tackles the computational bottleneck of simulating particle detector responses at high resolutions by introducing iCaloFlow, a framework using inductive normalizing flows and teacher-student distillation, achieving fast and high-fidelity simulation on detector geometries 10-100 times higher granularity than before.
Simulating particle detector response is the single most expensive step in the Large Hadron Collider computational pipeline. Recently it was shown that normalizing flows can accelerate this process while achieving unprecedented levels of accuracy, but scaling this approach up to higher resolutions relevant for future detector upgrades leads to prohibitive memory constraints. To overcome this problem, we introduce Inductive CaloFlow (iCaloFlow), a framework for fast detector simulation based on an inductive series of normalizing flows trained on the pattern of energy depositions in pairs of consecutive calorimeter layers. We further use a teacher-student distillation to increase sampling speed without loss of expressivity. As we demonstrate with Datasets 2 and 3 of the CaloChallenge2022, iCaloFlow can realize the potential of normalizing flows in performing fast, high-fidelity simulation on detector geometries that are ~ 10 - 100 times higher granularity than previously considered.