Understanding Event-Generation Networks via Uncertainties
Marco Bellagente, Manuel Haußmann, Michel Luchmann, Tilman Plehn
SciPost Phys. 13, 003 (2022) · published 21 July 2022
- doi: 10.21468/SciPostPhys.13.1.003
- Submissions/Reports
Abstract
Following the growing success of generative neural networks in LHC simulations, the crucial question is how to control the networks and assign uncertainties to their event output. We show how Bayesian normalizing flow or invertible networks capture uncertainties from the training and turn them into an uncertainty on the event weight. Fundamentally, the interplay between density and uncertainty estimates indicates that these networks learn functions in analogy to parameter fits rather than binned event counts.
Cited by 29
Authors / Affiliation: mappings to Contributors and Organizations
See all Organizations.
Funder for the research work leading to this publication