Speaker
Description
The small production cross-section of higgsinos poses a persistent challenge for detection at the LHC. We focus on a simplified R-parity violating supersymmetric model, where existing ATLAS limits on higgsino mass are around 450 GeV—significantly weaker than those for wino and bino counterparts. In this study, we investigate the potential to probe higgsino masses in the range of 400 to 1000 GeV. We employ a Graph Neural Network(GNN)-based LorentzNet classifier to tag boosted fat jets originating from top quarks and integrate data from tracker detectors to enhance classification accuracy.For each benchmark point, we trained separate Boosted Decision Trees (BDTs) in two mutually exclusive signal regions to distinguish signal from background effectively. By combining the statistical significance across both signal regions, we demonstrate that our methodology allows for probing higgsino masses up to 800 GeV at the high-energy, high-luminosity LHC.
| Field of contribution | Phenomenology |
|---|
