Home | Teaching |
Kihyuk (Ki) Hong | 홍기혁
Assistant Professor
Department of Industrial & Systems Engineering
KAIST (Korea Advanced Institute of Science & Technology)
✉️ Email
My primary area of research is artificial intelligence (AI) with a focus on reinforcement learning (RL). My main focus is on the design and analysis of computationally efficient algorithms with provable guarantees. I have worked on problems including non-stationary bandits, offline RL and constrained RL.
I was born and raised in South Korea. My alma maters are Johns Hopkins (BS, 2008), Stanford (MS, 2010) and University of Michigan (PhD, 2025. Advisor: Ambuj Tewari). Before starting my PhD, I worked on recommendation systems and search engines as a machine learning engineer at Meta (2010-2013) and Naver (2013-2018, 2019-2020).
[8] Marc Brooks*, Gabriel Durham*, Kihyuk Hong*, Ambuj Tewari. Generator-Mediated Bandits: Thompson Sampling for GenAI-Powered Adaptive Interventions. Preprint 2025. [paper]
[7] Kihyuk Hong, Ambuj Tewari. Offline Constrained Reinforcement Learning with Arbitrary Data Distributions under Partial Coverage. Preprint 2025. [paper]
[6] Kihyuk Hong, Ambuj Tewari. A Computationally Efficient Algorithm for Infinite-Horizon Average-Reward Linear MDPs. In Proceedings of the 42nd International Conference on Machine Learning, 2025*.* [paper]
[5] Kihyuk Hong, Woojin Chae, Yufan Zhang, Dabeen Lee, Ambuj Tewari. Reinforcement Learning for Infinite-Horizon Average-Reward Linear MDPs via Approximation by Discounted-Reward MDPs. In Proceedings of the 28th International Conference on Artificial Intelligence and Statistics, 2025. [paper]
[4] Woojin Chae, Kihyuk Hong, Yufan Zhang, Ambuj Tewari, Dabeen Lee. Learning Infinite-Horizon Average-Reward Linear Mixture MDPs of Bounded Span. In Proceedings of the 28th International Conference on Artificial Intelligence and Statistics, 2025. [paper]
[3] Kihyuk Hong, Ambuj Tewari. A Primal-Dual Algorithm for Offline Constrained Reinforcement Learning with Linear MDPs. In Proceedings of the 41st International Conference on Machine Learning, 2024*.* [paper]
[2] Kihyuk Hong, Yuhang Li, Ambuj Tewari. A Primal-Dual-Critic Algorithm for Offline Constrained Reinforcement Learning. In Proceedings of the 27th International Conference on Artificial Intelligence and Statistics, 2024. [paper]
[1] Kihyuk Hong, Yuhang Li, Ambuj Tewari. An Optimization-Based Algorithm for Non-Stationary Kernel Bandits without Prior Knowledge. In Proceedings of the 26th International Conference on Artificial Intelligence and Statistics, 2023. [paper]