Speaker
Description
We propose $w_i f_i$ ensembles, a novel framework to obtain asymptotic frequentist uncertainties on density ratios in the context of neural ratio estimation. In the case where the density ratio of interest is a likelihood ratio conditioned on parameters, for example a likelihood ratio of collider events conditioned on parameters of nature, it can be used to perform simulation-based inference on those parameters. We show how uncertainties on a density ratio can be estimated with ensembles and propagated to determine the resultant uncertainty on the estimated parameters. We then turn to an application in quantum chromodynamics (QCD), using ensembles to estimate the likelihood ratio between generated quark and gluon jets. We use this learned likelihood ratio to estimate the quark fraction in a mixed quark/gluon sample, showing that the resultant uncertainties empirically satisfy the desired coverage properties. We also examine the performance of this inferred likelihood ratio for reweighting gluon to quark jets, using the distributions of angularities as a benchmark.