Beomhan Baek

Incoming Ph.D. student at UCLA CS

Hello, my name is Beomhan Baek, and I am an incoming Ph.D. student at UCLA CS. I’m interested in developing rigorous and principled algorithms for machine learning. My research interests span a broad range of machine learning theory, including optimization, deep learning theory, and interactive decision-making.

Previously, I earned my B.S. degree in Mathematical Sciences from Seoul National University. During my undergraduate studies, I worked on optimization in deep learning, and was fortunate to be advised by Prof. Chulhee Yun.

News

  • 2026 I will join UCLA as a Ph.D. student in Computer Science.
  • 2025 Our paper on the implicit bias of per-sample Adam will appear at ICLR 2026.
  • 2025 Our paper on schedule-free methods for language model training will appear at NeurIPS 2025.

Selected publications

  1. 2026
    Implicit Bias of Per-sample Adam on Separable Data: Departure from the Full-batch Regime
    Beomhan Baek*, Minhak Song*, Chulhee Yun
    ICLR 2026
    NeurIPS 2025 Workshop on Optimization for Machine Learning
  2. 2025
    Through the River: Understanding the Benefit of Schedule-Free Methods for Language Model Training
    Minhak Song*, Beomhan Baek*, Kwangjun Ahn, Chulhee Yun
    NeurIPS 2025
    ICML 2025 Workshop on High-dimensional Learning Dynamics