Probing a Quarkophobic ${\mathbf{W}}^\prime$ at the High-Luminosity LHC via Vector Boson Fusion and Lorentz-Equivariant Point Cloud Learning

TL;DR

This paper explores detecting a quarkophobic W' boson at the High-Luminosity LHC using vector boson fusion and introduces a novel Lorentz-Equivariant Geometric Algebra Transformer for improved sensitivity in BSM searches.

Contribution

It presents a new phenomenological approach for W' boson detection and applies a novel Lorentz-equivariant point cloud learning method to enhance search sensitivity.

Findings

Significant improvement in signal sensitivity using the new method.

First application of point cloud learning in a BSM search.

Probing W' production via vector boson fusion at HL-LHC.

Abstract

The addition of a heavy charged vector gauge boson ${\mathbf{W}}^\prime$ to the Standard Model (SM) with negligible quark couplings ("quarkophobic") and triple gauge couplings can address issues with the SM, such as the B-meson anomalies and recent discrepancies in the W boson mass measurements. We present a phenomenology study probing ${\mathbf{W}}^\prime$ production through weak boson fusion in proton-proton collisions at the Large Hadron Collider. We operate under a simplified model with a large ${\mathbf{W}}^\prime$ decay width and consider final states with two jets, large missing transverse momentum, and one light lepton. Notably, we use point cloud learning for the first time in a BSM search$\unicode{x2014}$specifically, a novel Lorentz-Equivariant Geometric Algebra Transformer$\unicode{x2014}$providing significant improvement in signal sensitivity compared to traditional methods.