Hi, I'm Hyeong Jun - a Staff Researcher at LG Energy Solution. I work at the intersection of physics and machine learning, building models that are not just accurate, but also interpretable.
My focus is on prediction and diagnosis in real-world systems, where understanding why a model works matters as much as how well it works.
EMPLOYMENT
LG Energy Solution, Seoul, Korea
Staff Researcher, Battery Diagnosis and Prediction Team [2025.1 - present]
Korea Advanced Institute of Science and Technology (KAIST), Daejeon, Korea
Postdoctoral Researcher (MJ Han's group), Department of Physics [2022.4 - 2024.12]
Institute for Basic Science (IBS), Daejeon, Korea
Postdoctoral Researcher, Center for Theoretical Physics of Complex Systems [2018.3 - 2022.3]
Seoul National University (SNU), Seoul, Korea
Postdoctoral Fellow, Department of Physics and Astronomy (jointly in IBS-CCES) [2017.4 - 2018.2]
EDUCATION
Seoul National University, Seoul, Korea
PhD in Physics (2017), MSc in Physics (2012)
Sogang University, Seoul, Korea
BSc in Physics (2009)
more below
RESEARCH INTERESTS
Interpretable and physics-informed machine learning
Computational physics & optimization
Quantum materials & quantum information
COMPUTATION SKILLS
Machine Learning & Data Science
Interpretable, physics-informed ML for predictive modeling and anomaly detection
Frameworks: PyTorch, TensorFlow, scikit-learn
Languages: Python (≤3.13), Julia (≤1.11), C/C++ (≤C++17)
Tools: Git, OpenMP, MPI
Physics-Based Computational Methods
Dynamical Mean-Field Theory (DMFT) with Exact Diagonalization solver — main developer (Julia; link, C++)
DMFT + Magnetic Force Response Theory (MFT) — main developer (Julia;)
Density Matrix Renormalization Group (DMRG) — from scratch (C++); ITensor library
Hartree-Fock Mean-Field Theory — from scratch (C/C++)
Other Packages & Tools
DFT: VASP, WIEN2k, OpenMX
Quantum Monte Carlo: TRIQS, EDMFTF, ComCTQMC
Tensor network: ITensor (DMRG, TEBD)
PUBLICATIONS
