Speaker
Description
The study of longitudinally polarized vector bosons interactions in the high-energy regime provides direct insight into the electroweak symmetry breaking mechanism through the Goldstone Boson Equivalence Theorem. However, in vector boson scattering (VBS) diboson production, the fraction of longitudinally polarized bosons is small, of the order of 7%, making their measurement particularly challenging.
To date, polarization studies have focused on fully leptonic boson decay channels, where angular observables sensitive to polarization can be reconstructed. In this work, we explore the semi-leptonic VBS diboson final state as a benchmark, benefiting from a larger branching ratio for hadronic decays while retaining a clean experimental signature from the leptonically decaying boson. The larger statistics in this topology provide access to higher transverse-momentum regimes of the bosons.
We focus on events with a boosted hadronic vector boson reconstructed as a large-radius jet, where polarization information is not directly accessible experimentally. To address this limitation, we investigate the use of jet-level variables and constituents to recover polarization-sensitive observables, while the boson decaying leptonically is treated with a standard selection.
In this study, we perform a truth-level analysis using MadGraph+Pythia simulations, with object selections implemented to approximate detector acceptance. To evaluate the sensitivity to longitudinal polarization, we compare a baseline cut-based approach using classical jet and lepton observables with an analysis using machine learning-driven regression of polarization-sensitive variables.