Probing Singlet Vector-Like Top Quarks in the Hadronic tZ Channel at the HL-LHC using Machine and Deep Learning Architectures
Abstract
In this work, we study the single production of a vector-like singlet top partner T at the 14 TeV HL-LHC in the channel pp → T j with T → t Z, t → b W → b j j, and Z → νν̅. Signal and background samples are generated with MadGraph5\_aMC@NLO v3.5.11, showered with Pythia 8, and passed through Delphes. The dominant backgrounds are t t̅, t Z j, ZZ j j, and W Z j j (including charge conjugates). A hadronic pre-selection (Nj ≥ 3, Nb ≥ 1, Nℓ = 0) is imposed as trigger, followed by optimized kinematic cuts. We perform multivariate classification with Extreme Gradient Boosting (XGBoost) and a Graph Neural Network (GNN) based on jet-level features. Sensitivities at 3000 fb−1 are quoted using the Asimov significance, S / √(S + B), and an Asimov variant with a 20\% background systematic. The model parameters g* and RL are defined in Sec.~2, and a single global working point is used to avoid per-mass tuning bias. In the (g*, mT) scan, we present 2σ exclusion and 5σ discovery contours for RL = 0 and RL = 0.5. For RL = 0, 2σ exclusion corresponds to g* ∈ [0.17, 0.49] (0.16, 0.43) over mT ∈ [1.8, 2.7] TeV, while 5σ discovery corresponds to g* ∈ [0.27, 0.44] (0.26, 0.40) over mT ∈ [1.8, 2.2] TeV for XGBoost and GNN respectively. For RL = 0.5, the 2σ reach is g* ∈ [0.21, 0.48] (0.20, 0.43) over mT ∈ [1.8, 2.5] TeV, and the 5σ reach is g* ∈ [0.33, 0.43] (0.31, 0.49) over mT ∈ [1.8, 2.2] TeV, with the GNN yielding slightly stronger and smoother limits across the scan.
Related
- Probing the Energy Evolution and Hadronic Cascade Time Dependence of Forward-Backward Multiplicity Correlations in pp Collisions at √s = 0.9--$13.6$~TeV with AMPT
- Probing the Single Production of First-Generation Singlet Vector-like Leptons at Future e+e- Colliders
- Probing the singlet-triplet scotogenic model with lepton flavor violation, gravitational waves and dark matter
- Probing Collective and Individual Kondo Screening: Multi-Stage, Multi-Channel Kondo Effects in a C3-Symmetric Four-Impurity Model
- Understanding Energy Dependent Hadronic Calorimeter Response from a Machine Learning Perspective
BibTeX
@misc{sagheer2026probing,
title = {{Probing Singlet Vector-Like Top Quarks in the Hadronic tZ Channel at the HL-LHC using Machine and Deep Learning Architectures}},
author = {Haroon Sagheer and M. Tayyab Javaid and Ijaz Ahmed and Ali Hassan and Farzana Ahmad},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2606.03913},
eprint = {2606.03913},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2606.03913},
}
RIS
TY - UNPB
TI - Probing Singlet Vector-Like Top Quarks in the Hadronic tZ Channel at the HL-LHC using Machine and Deep Learning Architectures
AU - Haroon Sagheer
AU - M. Tayyab Javaid
AU - Ijaz Ahmed
AU - Ali Hassan
AU - Farzana Ahmad
PY - 2026
DA - 2026/10/11
PB - arXiv
DO - 10.48550/arxiv.2606.03913
UR - https://arxiv.org/abs/2606.03913
AB - In this work, we study the single production of a vector-like singlet top partner T at the 14 TeV HL-LHC in the channel pp → T j with T → t Z, t → b W → b j j, and Z → νν̅. Signal and background samples are generated with MadGraph5\_aMC@NLO v3.5.11, showered with Pythia 8, and passed through Delphes. The dominant backgrounds are t t̅, t Z j, ZZ j j, and W Z j j (including charge conjugates). A hadronic pre-selection (N_j ≥ 3, N_b ≥ 1, N_ℓ = 0) is imposed as trigger, followed by optimized kinematic cuts. We perform multivariate classification with Extreme Gradient Boosting (XGBoost) and a Graph Neural Network (GNN) based on jet-level features. Sensitivities at 3000 fb^−1 are quoted using the Asimov significance, S / √(S + B), and an Asimov variant with a 20\% background systematic. The model parameters g^* and R_L are defined in Sec.~2, and a single global working point is used to avoid per-mass tuning bias. In the (g^*, m_T) scan, we present 2σ exclusion and 5σ discovery contours for R_L = 0 and R_L = 0.5. For R_L = 0, 2σ exclusion corresponds to g^* ∈ [0.17, 0.49] (0.16, 0.43) over m_T ∈ [1.8, 2.7] TeV, while 5σ discovery corresponds to g^* ∈ [0.27, 0.44] (0.26, 0.40) over m_T ∈ [1.8, 2.2] TeV for XGBoost and GNN respectively. For R_L = 0.5, the 2σ reach is g^* ∈ [0.21, 0.48] (0.20, 0.43) over m_T ∈ [1.8, 2.5] TeV, and the 5σ reach is g^* ∈ [0.33, 0.43] (0.31, 0.49) over m_T ∈ [1.8, 2.2] TeV, with the GNN yielding slightly stronger and smoother limits across the scan.
ER -