Anomalous electroweak physics unraveled via evidential deep learning

The growth in beyond standard model (BSM) models and parametrizations has placed strong emphasis on systematically intercomparing within the range of possible models with controlled uncertainties. In this setting, the language of uncertainty quantification (UQ) provides quantitative metrics of assessing overlaps and discrepancies between models. We leverage recent machine learning (ML) developments in evidential deep learning (EDL) for UQ to separate data (aleatoric) and knowledge (epistemic) uncertainties in a model discrimination setting. In this study, we construct several potentially BSM-motivated scenarios for the anomalous electroweak interaction (AEWI) of neutrinos with nucleons in deep inelastic scattering ($\nu$DIS). These scenarios are then quantitatively mapped, as a demonstration, alongside Monte Carlo replicas of the CT18 PDFs used to calculate the $\Delta \chi^{2}$ statistic for a typical multi-GeV $\nu$DIS experiment, CDHSW. Our framework effectively highlights areas of model agreement and provides a classification of out-of-distribution (OOD) samples. By offering the opportunity to identify unexplored regions of parameter space while understanding domains of model overlap, the approach presented in this work can help facilitate efficient BSM model exploration and exclusion for future New Physics searches.