Jaewon Kim

Welcome! I am an undergraduate student in Semiconductor Engineering at Korea University and an undergraduate researcher at the Complex Data Analytics Lab (CDAL), advised by Prof. Junhee Seok.

My research interests center on advancing machine learning methods that are built on strong mathematical foundations while remaining applicable to real-world problems. I am particularly interested in how concepts from probability, differential equations, and physics can guide the design of more reliable and efficient learning algorithms, with applications ranging from generative models to reinforcement learning. I believe that blending these directions provides a pathway toward models that are not only powerful but also grounded in mathematical principles.

Research Interests

Generative Models
Stochastic Partial Differential Equations
Deep Reinforcement Learning
Physics-Informed Machine Learning

news

Jun 18, 2026	Our paper `Fokker–Planck Soft Actor–Critic` has been accepted to `IEEE TNNLS` (Early Access)
Mar 27, 2025	Started `mandatory military service` as a Social Service Agent, continuing research activities during off-hours.
Oct 23, 2024	Presented `OpticalGAN` at the 12th SK Hynix Academic Conference, receiving the `Outstanding Poster Award`
Feb 20, 2024	I started an `undergraudate research internship` at CDAL

selected publications

TNNLS
Fokker–Planck Soft Actor–Critic

Hyo-Seok Hwang, Jaewon Kim, and Junhee Seok

IEEE Transactions on Neural Networks and Learning Systems, 2026

Abs Bib HTML

Learning expressive and multimodal policies is essential for solving complex continuous control tasks. We propose FP-SAC, which reinterprets soft policy improvement through stochastic differential equations and derives a distribution-level objective from the corresponding Fokker–Planck equation. By minimizing the PDE residual with physics-informed learning and leveraging normalizing flows, FP-SAC enables stable and accurate learning of multimodal policies, achieving higher performance and greater stability on multigoal, MuJoCo, and Meta-World benchmarks.
@article{hwang2026fpsac, title = {Fokker--Planck Soft Actor--Critic}, author = {Hwang, Hyo-Seok and Kim, Jaewon and Seok, Junhee}, journal = {IEEE Transactions on Neural Networks and Learning Systems}, year = {2026}, publisher = {IEEE}, doi = {10.1109/TNNLS.2026.3699399}, }