Quartic Gauge-Higgs couplings: Constraints and Future Directions

TL;DR

This paper reviews current constraints on quartic Higgs-gauge couplings from LHC data, discusses theoretical limitations of existing methods, and explores future analysis improvements using advanced machine learning techniques.

Contribution

It introduces a more theoretically consistent framework for setting limits on quartic couplings and proposes the use of Graph Neural Networks to enhance sensitivity in future analyses.

Findings

Current LHC limits are robust at leading order.

Radiative corrections reveal limitations in the SM approach.

Graph Neural Networks can improve background discrimination.

Abstract

Constraints on quartic interactions of the Higgs boson with gauge bosons have been obtained by the experimental LHC collaborations focussing on the so-called $\kappa$ framework of flat rescalings of SM-like interactions in weak boson fusion (WBF) Higgs pair production. While such approaches are admissible to obtain a qualitative picture of consistency with the SM when the statistical yield is low, once more statistics become available a more theoretically consistent framework of limit setting is desirable. Reviewing the constraints provided at the Large Hadron Collider, we first show that these limits are robust when considered in a leading order context. Turning to radiative corrections, we demonstrate the limitations of this approach in the SM, and by adopting Higgs effective field theory techniques, we clarify the sensitivity from single Higgs measurements to rescalings of quartic Higgs-gauge couplings. We then discuss avenues for sensitivity improvements of WBF analyses employing Graph Neural Networks to combat the large contributing backgrounds.