Minhak Song
About Me
Hi! I am an undergraduate student at KAIST, double majoring in Industrial & Systems Engineering and Mathematical Sciences. My research interest lies in building the theoretical foundations of modern machine learning through the lens of optimization theory and statistics. Recently, I am focusing on understanding the training dynamics of gradient-based optimization methods in deep learning, closely working with Prof. Chulhee Yun. Feel free to reach out for discussions and potential collaborations!
Currently, I am on a leave of absence from KAIST to complete my mandatory military service in South Korea, which will last from February 2023 to November 2024.
Research Interests
- Machine Learning Theory
- Optimization
- Statistics
News
- [Jun. 2024] Our paper on SGD dynamics along Hessian eigenspaces is accepted to ICML 2024 Workshop on High-dimensional Learning Dynamics: The Emergence of Structure and Reasoning.
- [Jan. 2024] Our paper on the optimization characteristics of linear Transformers is accepted to ICLR 2024.
- [Oct. 2023] Our paper on the optimization characteristics of linear Transformers is accepted to NeurIPS 2023 Workshop on Mathematics of Modern Machine Learning as an oral presentation.
- [Sep. 2023] Our paper on understanding the Edge of Stability phenomenon in deep learning is accepted to NeurIPS 2023.
Publications
-
Manuscript, 2024.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.