Junu Choi
I am a Master’s student in Artificial Intelligence at Seoul National University, advised by Prof. Youngjoon Hong, studying generative modeling, probabilistic inference, and scientific machine learning.
Previously, I completed my B.S. in Computer Science and Mathematical Sciences at KAIST.
Research Interests
Connecting mathematical foundations in transport and diffusion with scientific discovery.
Papers
Efficient probabilistic weather forecasting at quarter-degree resolution with two-sided transport attention
Kantorovich Attention routes information from both sides (pull & push) so zero regions are left unread, closely matching the exact optimal transport plan. Trained on a single 8-GPU node, Astra has lower CRPS and ensemble-mean RMSE than ECMWF IFS ENS on all 600 variable–level–lead-time combinations, verified against ERA5.
Posterior Transport Matching for Inverse Problems
Posterior Transport Matching pairs numerical transport maps with minimum-energy stochastic rematching so each map’s inputs remain posterior-compatible using only a frozen denoiser and the observation operator.
