Probing Singlet Vector-Like Top Quarks in the Hadronic tZ Channel at the HL-LHC using Machine and Deep Learning Architectures

Abstract We study single production of a singlet vector-like top partner T at the 14 TeV HL-LHC in the channel pp → T j, with T → tZ, t → bjj and Z → ν ν̄, for mT = 1.8–3.1 TeV at the benchmark g ∗ = 0.4, RL = 0.5. After a hadronic preselection and a χ2 -based reconstruction of the hadronic top quark, signal and background are separated with an XGBoost classifier and a graph neural network (GNN). Both classifiers receive identical event information and are trained on common process-stratified partitions with three random seeds at every mass. Their test AUCs lie in the range 0.984–0.993 and differ by at most 0.0023, so the two architectures perform equivalently once the input information is matched. The background model comprises eighteen Standard Model processes, including Z(ν ν̄)+jets, lost-lepton top-quark, W +jets and tt̄V production, and the high-score background tail is evaluated with simulated samples that are statistically independent of the training. Taking the total background at its 95% confidence-level upper bound, the expected discovery significance at 3000 fb−1 is 3.16–4.23 at mT = 1.8 TeV, and three-standard-deviation evidence and 95% confidence-level exclusion are reached up to mT ≃ 1.82 TeV and mT ≃ 2.13 TeV, respectively, with both classifiers. The small signal-to-background ratio in the selected region makes a data-driven constraint of the background normalization essential for realizing this reach. Because the benchmark exceeds the bounds from CKM unitarity, t → uZ decays and electroweak precision data in the minimal model, we also give expected limits on σ(pp → T j) B(T → tZ), which lie between 41 and 64 fb for mT = 1.8–3.1 TeV and correspond to g ∗ ≃ 0.28 at 1.8 TeV for RL = 0.5. Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. Article funded by SCOAP3 and published under licence by Chinese Physical Society and the Institute of High Energy Physics of the Chinese Academy of Science and the Institute of Modern Physics of the Chinese Academy of Sciences and IOP Publishing Ltd.

Authors

Institutions

Publication Details

Journal
Chinese Physics C
Published
2026-11-17
DOI
https://doi.org/10.1088/1674-1137/aec2cd
Citations
1
Primary Topic
Particle physics theoretical and experimental studies
Type
article
Field-Weighted Citation Impact
0.64

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Probing Singlet Vector-Like Top Quarks in the Hadronic tZ Channel at the HL-LHC using Machine and Deep Learning Architectures

Farzana Ahmad, Haroon Sagheer, M. Tayyab Javaid, Ali Hassan et al.
1 citations
Chinese Physics C
Particle physics theoretical and experimental studies
0.64
article

Probing Singlet Vector-Like Top Quarks in the Hadronic tZ Channel at the HL-LHC using Machine and Deep Learning Architectures

Farzana Ahmad, Haroon Sagheer, M. Tayyab Javaid, Ali Hassan, Ijaz Ahmed
article en
1 citations

Abstract

Abstract We study single production of a singlet vector-like top partner T at the 14 TeV HL-LHC in the channel pp → T j, with T → tZ, t → bjj and Z → ν ν̄, for mT = 1.8–3.1 TeV at the benchmark g ∗ = 0.4, RL = 0.5. After a hadronic preselection and a χ2 -based reconstruction of the hadronic top quark, signal and background are separated with an XGBoost classifier and a graph neural network (GNN). Both classifiers receive identical event information and are trained on common process-stratified partitions with three random seeds at every mass. Their test AUCs lie in the range 0.984–0.993 and differ by at most 0.0023, so the two architectures perform equivalently once the input information is matched. The background model comprises eighteen Standard Model processes, including Z(ν ν̄)+jets, lost-lepton top-quark, W +jets and tt̄V production, and the high-score background tail is evaluated with simulated samples that are statistically independent of the training. Taking the total background at its 95% confidence-level upper bound, the expected discovery significance at 3000 fb−1 is 3.16–4.23 at mT = 1.8 TeV, and three-standard-deviation evidence and 95% confidence-level exclusion are reached up to mT ≃ 1.82 TeV and mT ≃ 2.13 TeV, respectively, with both classifiers. The small signal-to-background ratio in the selected region makes a data-driven constraint of the background normalization essential for realizing this reach. Because the benchmark exceeds the bounds from CKM unitarity, t → uZ decays and electroweak precision data in the minimal model, we also give expected limits on σ(pp → T j) B(T → tZ), which lie between 41 and 64 fb for mT = 1.8–3.1 TeV and correspond to g ∗ ≃ 0.28 at 1.8 TeV for RL = 0.5. Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. Article funded by SCOAP3 and published under licence by Chinese Physical Society and the Institute of High Energy Physics of the Chinese Academy of Science and the Institute of Modern Physics of the Chinese Academy of Sciences and IOP Publishing Ltd.

Chinese Physics C
University of Seoul (KR), Riphah International University (PK), University of Catania (IT), Seoul Medical Center (KR), The Seoul Institute (KR), University College of Islamabad (PK)
National Centre for Physics
Openalex Percentile: Top 35%
Particle physics theoretical and experimental studies
0.64
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon