Anomaly detection with flow-based fast calorimeter simulators

TL;DR

This paper demonstrates that flow-based calorimeter simulators can be used for unsupervised anomaly detection by leveraging the likelihood scores, providing a complementary approach to supervised classifiers in identifying unusual electromagnetic showers.

Contribution

It introduces a novel use of normalizing flow models for unsupervised anomaly detection in calorimeter simulations without additional training, enhancing detection capabilities.

Findings

Likelihood scores effectively identify unlikely showers

Method struggles with nearly collinear signal photons

Combining supervised and unsupervised methods improves detection

Abstract

Recently, several normalizing flow-based deep generative models have been proposed to accelerate the simulation of calorimeter showers. Using CaloFlow as an example, we show that these models can simultaneously perform unsupervised anomaly detection with no additional training cost. As a demonstration, we consider electromagnetic showers initiated by one (background) or multiple (signal) photons. The CaloFlow model is designed to generate single photon showers, but it also provides access to the shower likelihood. We use this likelihood as an anomaly score and study the showers tagged as being unlikely. As expected, the tagger struggles when the signal photons are nearly collinear, but is otherwise effective. This approach is complementary to a supervised classifier trained on only specific signal models using the same low-level calorimeter inputs. While the supervised classifier is also highly effective at unseen signal models, the unsupervised method is more sensitive in certain regions and thus we expect that the ultimate performance will require a combination of these approaches.