Artigo Acesso aberto Revisado por pares

Understanding Event-Generation Networks via Uncertainties

2022; SciPost.org; Volume: 13; Issue: 1 Linguagem: Inglês

10.21468/scipostphys.13.1.003

ISSN

2542-4653

Autores

Marco Bellagente, Manuel Haußmann, Michel Luchmann, Tilman Plehn,

Tópico(s)

High-Energy Particle Collisions Research

Resumo

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.

Referência(s)