Yeseul Cho
About Me
Hi! I am an AI Engineer Intern at Nota AI. I received my master’s degree from the Kim Jaechul Graduate School of AI at KAIST, where I was fortunate to be advised by Chulhee Yun.
My research interests broadly lie in understanding and improving modern neural networks, with a particular focus on large language models. My previous work has explored data selection and pruning, as well as the adaptation and modification of trained models through problems such as model merging and machine unlearning.
Going forward, I am particularly interested in understanding the optimization dynamics that arise during large language model training, including how data, optimization algorithms, and training scale interact. I am also interested in optimizing models for inference, with the broader goal of improving how large-scale models are trained and deployed.
Research Interests
- Data-Centric Learning
- Optimization & Training Dynamics
- Inference Optimization
Education
- M.S. in Artificial Intelligence, Korea Advanced Institute of Science and Technology, Sep. 2024 - Aug. 2026
- B.S. in Biomedical Engineering, Korea University, Mar. 2020 - Aug. 2024
News
- [Sep. 2026] Our paper about machine unlearning is accepted to the NeurIPS 2026 OPT Workshop!
- [Sep. 2026] Our papers about label-efficient dataset pruning and Muon optimization are accepted to NeurIPS 2026!
- [Aug. 2026] I joined Nota AI as an AI Engineer Intern.
- [Aug. 2026] I earned my M.S. in Artificial Intelligence from KAIST.
- [May 2025] Our paper about lightweight dataset pruning is accepted to ICML 2025!
Publications
(* denotes equal contribution)
-
Machine Unlearning with a Destination: Tracking a Minimizer Path to the Retain-Only ObjectiveNeurIPS 2026 Workshop on Optimization for Machine Learning (OPT)
-
NeurIPS 2026ICML 2026 Workshop on High-dimensional Learning Dynamics (HiLD)
-
ICML 2025ICLR 2025 Workshop on Navigating and Addressing Data Problems for Foundation Models (DataFM)