Bio
Pei Zhang is a computational scientist in the Multiscale Materials Group at ORNL. She is broadly interested in solving multiscale multiphysics problems using data-driven and physics-based modeling approaches. Her recent work focuses on graph neural networks (GNNs) and large language models (LLMs) for molecular properties, efficient reinforcement learning (RL) algorithm development, and dimension reduction for stiff chemical dynamic systems.
Pei joined ORNL in 2019 as a postdoctoral research associate. Her postdoctoral work includes developing reduced-order models for chemical kinetics using deep neural networks, uncertainty quantification for robust machine learning models, and a Julia software prototype for solving reactive transport with automatic differentiation.
Prior to joining ORNL, Pei received her PhD degree in Aerospace Engineering from Purdue University, where she was awarded Bilsland Dissertation Fellowship, Koerner Scholarship, and NSF student INTERN funding. Her dissertation topic is turbulent combustion modeling using large eddy simulation (LES) and transported probability density function (TPDF) methods.