Jihwan Kim

prof_jh.jpg

I am an undergraduate at Seoul National University studying mathematics and computer science. My research focuses on the theory and design of optimization algorithms for machine learning, including scaling laws, implicit bias, normalization, and preconditioning. I am particularly interested in understanding how optimizer dynamics and data geometry shape learned solutions and downstream performance.

news

May 01, 2026 Our paper on Ski Rental with Distributional Advice is accepted at ICML 2026.
Jan 26, 2026 Our paper on Scaling Law of SignSGD is accepted at ICLR 2026.

selected publications

  1. ICML
    Robust and Consistent Ski Rental with Distributional Advice
    Jihwan Kim and Chenglin Fan
    In Forty-Third International Conference on Machine Learning, 2026
  2. ICLR
    Scaling Laws of SignSGD in Linear Regression: When Does It Outperform SGD?
    Jihwan Kim, Dogyoon Song, and Chulhee Yun
    In The Fourteenth International Conference on Learning Representations, 2026
  3. CiC
    OverModRaise: Reducing Modulus Consumption of CKKS Bootstrapping
    Jihwan Kim, Jung Hee Cheon, and Yongdong Yeo
    IACR Communications in Cryptology, 2025