Working Paper
(NEW! ) Same Concept, Different Directions: Cross-Modal Feature Heterogeneity in Sparse Autoencoders
Chungpa Lee*, Jihoon Kwon*, Kyle Min, Jy-yong Sohn (*:Â equal contribution)
Mechanistic Interpretability Workshop at the Forty-Third International Conference on Machine Learning (ICML Workshop), 2026Poisson Regression with Additive Exponential Mean: Statistical Modeling and Insurance Applications
Chungpa Lee, Joseph H.T. Kim
(Under Review)
Publications
2026
(NEW! ) How to Correctly Report LLM-as-a-Judge Evaluations [arXiv] [github]
Chungpa Lee, Thomas Zeng, Jongwon Jeong, Jy-yong Sohn, Kangwook Lee
Forty-Third International Conference on Machine Learning (ICML), 2026(NEW! ) Fine-Tuning Without Forgetting In-Context Learning: A Theoretical Analysis of Linear Attention Models [arXiv] [github]
Chungpa Lee, Jy-yong Sohn, Kangwook Lee
Forty-Third International Conference on Machine Learning (ICML), 2026Transformers in the Dark: Navigating Unknown Search Spaces via Bandit Feedback [paper] [arXiv]
Jungtaek Kim, Thomas Zeng, Ziqian Lin, Minjae Lee, Chungpa Lee, Jy-yong Sohn, Hyung Il Koo, and Kangwook Lee
Transactions on Machine Learning Research (TMLR), 2026
2025
On the Similarities of Embeddings in Contrastive Learning [paper] [arXiv] [github]
Chungpa Lee, Sehee Lim, Kibok Lee, Jy-yong Sohn
Forty-Second International Conference on Machine Learning (ICML), 2025A Theoretical Framework for Preventing Class Collapse in Supervised Contrastive Learning [paper] [arXiv] [github]
Chungpa Lee, Jeongheon Oh, Kibok Lee, Jy-yong Sohn
The 28th International Conference on Artificial Intelligence and Statistics (AISTATS), 2025A Generalized Theory of Mixup for Structure-Preserving Synthetic Data [paper] [arXiv] [github]
Chungpa Lee, Jongho Im, Joseph H.T. Kim
The 28th International Conference on Artificial Intelligence and Statistics (AISTATS), 2025
2024