Jiseok Chae

prof_pic.jpg

E7, #2102

291 Daehak-ro, Yuseong-gu

Daejeon, Republic of Korea

I am a postdoctoral researcher, formally appointed at the Natural Science Research Institute at KAIST, but affiliated with and mainly working with the PDE & AI Lab at Yonsei University, led by Prof. Dohyun Kwon. I received my Ph.D. in Mathematical Sciences from KAIST under the supervision of Prof. Donghwan Kim. Before that, I completed my bachelor’s degree in Mathematical Sciences from KAIST, with a minor in Computer Science.

I study optimization theory and algorithms for machine learning. Recently, I also got interested in how optimization methods work in real world machine learning applications (implicit biases, generalization properties, etc.). If you are interested in collaborating with me, or if you want to know more about anything that is on this website, feel free to contact me by email.

News

Feb 10, 2026 I received the Outstanding Paper Award from KAIST for my doctoral dissertation, Two Theoretical Advances in Generative Models: minimax optimization methods with improved convergence and provable benign overfitting.
Jun 9, 2025 One paper got accepted to the 3rd Workshop on High-dimensional Learning Dynamics, in ICML 2025.
Apr 29, 2025 I will be presenting at ICCOPT 2025. See you in Los Angeles!

Selected publications

2024

  1. NeurIPS
    Stochastic Extragradient with Flip-Flop Shuffling & Anchoring: Provable Improvements
    Jiseok Chae, Chulhee Yun, and Donghwan Kim
    The Thirty-eighth Annual Conference on Neural Information Processing Systems, 2024
  2. ICLR
    Two-timescale Extragradient for Finding Local Minimax Points
    Jiseok Chae, Kyuwon Kim, and Donghwan Kim
    The Twelfth International Conference on Learning Representations, 2024