Ephemeral Learning -- Augmenting Triggers with Online-Trained Normalizing Flows

Abstract

The large data rates at the LHC require an online trigger system to select relevant collisions. Rather than compressing individual events, we propose to compress an entire data set at once. We use a normalizing flow as a deep generative model to learn the probability density of the data online. The events are then represented by the generative neural network and can be inspected offline for anomalies or used for other analysis purposes. We demonstrate our new approach for a toy model and a correlation-enhanced bump hunt.

17 pages, 9 figures, minor changes to text, addressed referee comments

References in corpus (8)

• The CMS trigger system
• MADE: Masked Autoencoder for Distribution Estimation
• Classification without labels: Learning from mixed samples in high energy physics
• Extending the Bump Hunt with Machine Learning
• On hypothesis testing, trials factor, hypertests and the BumpHunter
• Targeting Multi-Loop Integrals with Neural Networks
• A comprehensive real-time analysis model at the LHCb experiment
• Practical Lossless Compression with Latent Variables using Bits Back Coding

Cited by in corpus (12)

• Machine Learning and LHC Event Generation
• MadNIS -- Neural Multi-Channel Importance Sampling
• The MadNIS Reloaded
• Anomaly Detection under Coordinate Transformations
• CaloFlow for CaloChallenge Dataset 1
• Calorimeter shower superresolution
• Differentiable MadNIS-Lite
• Anomaly detection with flow-based fast calorimeter simulators
• TopicFlow: Disentangling quark and gluon jets with normalizing flows
• Unifying Simulation and Inference with Normalizing Flows
• Foundations of automatic feature extraction at LHC--point clouds and graphs
• Observable Optimization for Precision Theory: Machine Learning Energy Correlators