About

Hi! I recently graduated from Seoul National University with a double major in Computer Science and Engineering and Mathematical Sciences. My primary interests lie in reinforcement learning theory and, more broadly, decision-making under uncertainty, including areas such as multi-agent learning and game theory.

Here is my CV - Curriculum Vitae.
While CV contains my academic or other offical records, you can check out my Somewhat Unofficial Records and Facts at the link.

News

I am delighted to announce that two of my papers, Lasso Bandit with Compatibility Condition on Optimal Arm and Minimax Optimal Reinforcement Learning with Quasi-Optimism, have been accepted at ICLR 2025! Details will be updated soon.
[OpenReview link - Lasso Bandit with Compatibility Condition on Optimal Arm]
[OpenReview link - Minimax Optimal Reinforcement Learning with Quasi-Optimism]

Experiences

I am fortunate to work as an undergraduate researcher under the supervision of Professor Min-hwan Oh at Seoul National University. So far, I have written three papers: one is soon to be published, and the other two are currently under review for a conference.

In addition, I have actively participated in competitive programming. I competed in the 47th ICPC World Finals in Luxor, Egypt, in April 2024 and won a Silver Medal (7th place) and was named Asia Pacific Champion, which was an incredibly exciting experience.

Publications

[1] Harin Lee and Min-hwan. Oh, “Minimax optimal reinforcement learning with quasi-optimism”, International Conference on Learning Representations, 2025, (To appear). [arxiv]
[2] Harin Lee, Taehyun. Hwang, and Min-hwan Oh, “Lasso bandit with compatibility condition on optimal arm”, International Conference on Learning Representations, 2025, (To appear). [arxiv]
[3] Harin Lee and Min-hwan. Oh, “Improved regret of linear ensemble sampling”, Advances in Neural Information Processing Systems, vol. 37, 2025. [arxiv] [conf]

Blog

I write blog posts on noteworthy mathematical topics. I am also planning to share my thoughts on reinforcement learning and learning theory. I might add commentaries on my papers, though most of the time the papers should explain themselves sufficiently.