Byoungwoo Park

Visiting Scholar @ Georgia Tech | PhD student @ KAIST

I am an PhD student at KAIST advised by Juho Lee. I am also fortunate to be a visiting scholar at Georgia Tech, working closely with Yongxin Chen, and being mentored by Guan-Horng Liu.

My primary research interests are in GenAI and AI4Science, focusing on developing theoretically sound yet practical methods inspired by dynamic optimal transport, stochastic optimal control. My research covers unpaired image translation, time series modeling, fMRI representation learning, single-cell RNA sequencing, and transition path sampling.

Contact: bw.park [at] kaist [dot] ac [dot] kr
Follow: Google Scholar | LinkedIn | bw-park | 𝕏 @bw__park

news

[04/2026] FAS accepted to ICML 2026.

[01/2026] BDO accepted to ICLR 2026.

[09/2025] FAS and Self-distilled TSMC accepted to FPI@NeurIPS 2025 and received the AITHYRA Travel Grant.

[02/2025] ACSSM accepted to ICLR 2025 (Oral 1.8%)

[10/2024] DBFS accepted to NeurIPS 2024 and received the NeurIPS 2024 Scholar Award.

selected publications

(*) denotes equal contribution, (^‡) denotes equal advising. See publications for the full list.

Functional Adjoint Sampler: Scalable Sampling on Infinite Dimensional Spaces

Byoungwoo Park, Juho Lee, and Guan-Horng Liu

International Conference on Machine Learning (ICML), 2026
NeurIPS 2025 Workshop on Frontiers in Probabilistic Inference

PDF
Multi-Marginal Schrödinger Bridge Matching

Byoungwoo Park, and Juho Lee

Preprint

PDF Code
Amortized Control of Continuous State Space Feynman-Kac Model for Irregular Time Series

Byoungwoo Park, Hyungi Lee, and Juho Lee

International Conference on Learning Representations (ICLR), 2025
Oral Presentation at ICLR 2025 [Top 1.8%]
NeurIPS 2024 Workshop on Bayesian Decision-making and Uncertainty

PDF Code
Stochastic Optimal Control for Diffusion Bridges in Function Spaces

Byoungwoo Park, Jungwon Choi, Sungbin Lim^‡, and Juho Lee^‡

Advances in Neural Information Processing Systems (NeurIPS), 2024

PDF Code