The 2012 discovery of the Higgs boson completed the Standard Model, yet compelling evidence — both experimental and theoretical — points to physics beyond it. My research spans three interconnected fronts: searching for new heavy particles in multi-lepton final states at the LHC, making precision measurements of Higgs boson properties, and developing quantum machine learning algorithms to address the unprecedented computational challenges of next-generation collider experiments.
Searching for heavy resonances decaying to four leptons in association with missing transverse energy or jets using the full ATLAS Run-II dataset. I designed the first ATLAS search in this channel, exploring 2HDM+S and baryogenesis scenarios across 320–1300 GeV mass ranges.
Precision measurement of the Higgs boson total width and coupling constraints via off-shell ZZ production, and searches for non-resonant Higgs pair production in four-lepton final states. Focused on the ZZ → 4ℓ and ZZ → 2ℓ2ν combined channels.
Developing quantum support-vector machines and quantum transformer architectures for jet classification and event selection in HEP. Benchmarking performance on IBM and Origin Quantum hardware for future collider applications at the CEPC and LHC.
Developing the first software-based neutral hadron tagger for the BESIII experiment, designed to compensate for the absence of a dedicated hadronic calorimeter. The tagger uses a transformer-based deep neural network trained on EMC shower shape variables and TOF timing information to classify K⁰L, neutrons, and antineutrons on an event-by-event basis.
The full workflow spans training and validation in Python, export of the trained model to ONNX format, and integration into a dedicated BOSS (BESIII Offline Software System) algorithm that reads the ONNX file at runtime and applies the model weights directly within the standard BESIII reconstruction chain — making the tagger immediately available to physics analyses across the collaboration. Preliminary results achieve AUC > 0.95 across all hadron classes.
Higgs CP property measurements and differential cross-section studies for the proposed Circular Electron Positron Collider. Developed Monte Carlo generator frameworks for e⁺e⁻ → ZH → ℓ⁺ℓ⁻H processes under multiple CP-coupling assumptions.
Research Vision
A central challenge in experimental particle physics is that many of the most interesting phenomena are only indirectly observable. Neutral hadrons, long-lived particles, and weakly interacting states often leave incomplete or diffuse signatures in detectors, limiting the reach of traditional reconstruction methods.
My research aims to address this challenge by developing advanced inference techniques that connect detector-level signals to underlying physical processes. By combining detector physics, statistical methods, and modern machine learning, I work to recover information that would otherwise be inaccessible.
This approach enables both precision measurements — such as constraints on Higgs boson properties — and searches for new physics in challenging final states. In the long term, I aim to contribute to a paradigm in which reconstruction is not strictly limited by detector hardware, but can be extended through adaptive, data-driven methodologies that enhance the physics potential of existing and future experiments.