Physics
Tidbits about my professional work and perspectives
Physics-Informed ML for Quantum Magnets
CurrentPredicting high-Tₓ superconductors, universal electron-phonon coupling, and magnetic transition temperatures using physics-informed machine learning.
Projected Density Matrix Sampling
PublishedA continuous-time quantum Monte Carlo for the low-lying spectrum of generic lattice Hamiltonians, free of Trotter errors. Physical Review D 113, 074511 (2026).
Quantum Materials Synthesis
Undergrad ResearchSynthesis and single-crystal growth of frustrated spinels, perovskites, and borides for neutron scattering studies.
Atmospheric Pollution Modeling
PublishedHYSPLIT trajectory modeling of transboundary PM₂.₅ transport in Nepal and the Indo-Gangetic Plain. Atmospheric Pollution Research (2026).
HL-LHC Track Reconstruction
Undergrad ResearchWedge and spacepoint algorithms for real-time silicon track-trigger reconstruction at the HL-LHC, targeting metastable dark matter candidates.
Short posts on quantum condensed matter, numerical methods, and ideas I find worth thinking about.
- Soon
Why the Sign Problem is Hard (and Why We Try Anyway)
An informal look at the sign problem in quantum Monte Carlo and the strategies used to work around it.
- Soon
What is a Frustrated Magnet?
Spins on a triangle, competing interactions, and the strange phases that emerge when a system cannot satisfy all its constraints simultaneously.
- Soon
Machine Learning Meets Quantum Materials
How physics-informed ML is changing the way we search for high-Tc superconductors and novel magnetic materials.
Lecture notes and expository PDFs on topics I have worked through carefully enough to write down.
- Soon
Notes on the Shastry–Sutherland Model
Derivation of the orthogonal-dimer ground state, exact solution, and overview of the phase diagram.
- Soon
Introduction to Quantum Monte Carlo
Path integral formulation, importance sampling, and the PDMS method.
- Soon
Tensor Networks: A Physicist's Introduction
MPS, DMRG, and why contracting a tensor network is hard in general but tractable in 1D.
Physics-Informed ML for Quantum Magnets
My current research develops physics-informed machine learning methods for quantum magnetic materials. The three main threads are: predicting high-Tₓ superconductors from material descriptors; building a universal electron-phonon coupling calculator; and large-scale screening for materials with high Curie and Néel temperatures.
Projected Density Matrix Sampling for Lattice Hamiltonians
Quantum Monte Carlo methods are among the most powerful tools for studying quantum many-body systems, but frustrated models generically suffer from the sign problem. Even where the ground state is accessible, obtaining excited states reliably remains an open challenge.
I developed the Projected Density Matrix Sampling (PDMS) method: a continuous-time path-integral Monte Carlo approach for computing low-lying spectra of generic lattice Hamiltonians. PDMS samples within a carefully chosen projection subspace while maintaining ergodicity and detailed balance, achieving rapid convergence in sign-free systems and yielding spectral information even in the presence of sign problems. Applied to the Shastry–Sutherland model and the 1D Transverse Field Ising Model near the E₈ critical point. First-author paper under review at Physical Review D.
Quantum Materials Synthesis and Characterization
I joined the Haravifard Lab in my first year to work on frustrated quantum materials. For spinel compounds (Cr-doped ZnAl₂O₄ and Li-doped ZnV₂O₄), I synthesized phase-pure samples via solid-state techniques and used high-resolution XRD with Rietveld refinement to determine lattice parameters across dopings and quantify deviations from Vegard's law. I also performed arc melting on TmB₄ and HoB₄, metallic altermagnet candidates related to the Shastry–Sutherland lattice.
For perovskites, I synthesized phase-pure NdFeO₃, a theoretical topological magnon candidate, and grew a single crystal using the optical floating zone furnace. I performed Laue scattering for orientation alignment and PPMS measurements to obtain temperature-dependent magnetization and hysteresis curves. The crystal is currently being re-grown for inelastic neutron scattering studies.
Atmospheric Pollution and Climate Physics
Working with Prof. McAdoo's group, I modeled transboundary PM₂.₅ dynamics in Nepal and the Indo-Gangetic Plain using HYSPLIT back-trajectory analysis, aerosol optical depth (AOD), and potential source contribution functions (PSCF). I examined diurnal PM₂.₅ patterns, planetary boundary layer height, and ventilation coefficients to isolate local and non-local pollution sources. Co-first-author paper published in Atmospheric Pollution Research, 2026.
Collider Physics for HL-LHC Upgrade
As a Huang Fellow (2023), I designed track reconstruction algorithms for the HL-LHC upgrade with Prof. Kotwal's group. I implemented a wedge generation algorithm that segments concentric silicon detectors and a spacepoint divider that assigns collision data to those wedges, enabling parallelized real-time track-trigger reconstruction for metastable dark matter candidates. Contributions acknowledged in a published paper in Scientific Reports.