Jaehoon Hahm

Hi! I am a Physics Ph.D. candidate at the University of Illinois Urbana-Champaign, advised by Bryan K. Clark. With Bryan, I study quantum many-body systems using neural quantum states, quantum Monte Carlo, and generative methods for learning quantum states from classical shadows. I also collaborate with Romit R. Choudhury on compositional alignment and reward steering for diffusion models, and with Jaewoong Choi on reward steering of one-step generators.

Previously, I earned an M.S. in Data Science at Seoul National University, advised by Joonseok Lee, and a B.S. in Physics with a minor in Mathematics there.

Research interests

• Geometric deep learning and generative modeling for AI for science.
• Optimal transport and geometry-aware diffusion and flow models.
• Discrete flow matching for materials and molecular structures.
• Neural quantum states, quantum Monte Carlo, and subspace diagonalization.

At Meta Reality Labs, I am a Research Scientist Intern working with Jogendra N. Kundu on self-distillation, self-supervised learning, and representation learning for AR/VR devices, including eye tracking with limited target-device data.

Links: Google Scholar · LinkedIn · Email · CV