Minhak Song
About Me
Hi! My name is Minhak Song (송민학), and I am an undergraduate student at KAIST, majoring in Mathematical Sciences (minor in Industrial and Systems Engineering). My research interests lie in the theoretical foundations of modern machine learning, with the goal of bridging theory and practice. Recently, I have been studying the training dynamics of optimization algorithms in deep learning under the guidance of Prof. Chulhee Yun. Currently, I am a visiting student at the University of Washington, hosted by Prof. Simon Du, where I am working on reinforcement learning from human feedback (RLHF) from an optimization perspective.
Feel free to reach out for discussions and potential collaborations!
Research Interests
- Machine Learning Theory
- Optimization
- Statistics
News
- [May. 2025] Our paper on how the dataset, network architecture, and optimizer influence progressive sharpening is accepted to ICML 2025.
- [Jan. 2025] Our paper on SGD dynamics along Hessian eigenspaces is accepted to ICLR 2025.
- [Jan. 2025] I joined Prof. Simon Du’s group as a visiting student at the Paul G. Allen School of Computer Science & Engineering at the University of Washington.
- [Jan. 2024] Our paper on the optimization characteristics of linear Transformers is accepted to ICLR 2024.
- [Sep. 2023] Our paper on understanding the Edge of Stability phenomenon in deep learning is accepted to NeurIPS 2023.
Publications
-
International Conference on Machine Learning (ICML), 2025.
-
International Conference on Learning Representations (ICLR), 2025.ICML 2024 Workshop on High-dimensional Learning Dynamics: The Emergence of Structure and Reasoning.
-
International Conference on Learning Representations (ICLR), 2024.NeurIPS 2023 Workshop on Mathematics of Modern Machine Learning (Oral).
-
Neural Information Processing Systems (NeurIPS), 2023.