Starting in June 2023, Weishi Wang worked as a Machine Learning Physics Graduate Student Intern at Lawrence Livermore National Laboratory. During his internship, Weishi focused on the design of physics-inspired machine learning interatomic potential (MLIP) models for large-scale atomistic simulation. Compared to conventional ab initio methods such as density functional theory and multiconfigurational wavefunction methods, MLIP provides a new way to learn the high-dimensional potential energy surface of electronic structure through reference data sets. By encoding physical intuitions of atomic interactions into the architectural design of MLIP models, Weishi helped bridge the gap between first-principle characterizations of quantum many-body systems and data-driven modeling techniques.
Dartmouth | Department of Physics and Astronomy