Kun Wang
Department of Computer Science
Purdue University
[Google scholar][Linkedin][Email]
About me
Welcome to my homepage! I am a Ph.D. candidate in the Department of Computer Science at Purdue University, where I am fortunate to be advised by Prof. Paul Valiant and Prof. Steve Hanneke.
My research focuses on sequential decision-making under uncertainty, spanning multi-armed bandits, reinforcement learning, and their applications to large language models and recommendation systems. At its core, my work asks how to make reliable decisions in unknown environments by balancing exploration and exploitation, optimizing for long-term outcomes rather than immediate rewards. These questions bear directly on exploration strategies in modern recommendation systems, the design of LLM-based agents, and the post-training of large language models.
In summer 2026, I was a Ph.D. Software Engineering Intern at Google, working on YouTube Shorts recommendations.
Education
- Ph.D. in Computer Science, Purdue University, 2022 - present
- M.Sc. in Computer Science, Shanghai Jiao Tong University, 2019 - 2022
- B.Eng. in Information Security, Xidian University, 2015 - 2019
Publications
Note: (α-β) denotes alphabetical author order; * denotes equal contribution.
- (α-β)Jasper C.H. Lee, Suraj Mathashery, Paul Valiant, Kun Wang
Optimal Sub-Gaussian Mean Estimation in Low Dimensions.
- (α-β)Steve Hanneke, Farnam Mansouri, Alireza F. Pour, Amirreza Shaeiri, Kun Wang
PAC Learning with Multiple Correct Answers (arXiv).
- Kun Wang
Optimal No-Regret Learning for Repeated Prophet Inequality (arXiv).
- (α-β)Steve Hanneke, Kun Wang
A Complete Characterization of Learnability for Adversarial Noisy Bandits (arXiv).
- Junyan Liu*, Ziyun Chen*, Kun Wang*, Haipeng Luo, Lillian J. Ratliff
Improved Regret and Contextual Linear Extension for Pandora's Box and Prophet Inequality (NeurIPS 2025).
- (α-β)Steve Hanneke, Kun Wang
A Complete Characterization of Learnability for Stochastic Noisy Bandits (ALT 2025).
- (α-β)Simina Brânzei, MohammadTaghi Hajiaghayi, Reed Phillips, Suho Shin, Kun Wang
Dueling over Dessert, Mastering the Art of Repeated Cake Cutting (NeurIPS 2024).
- Kun Wang, Jing Dong, Baoxiang Wang, Shuai Li
Cascading Bandit under Differential Privacy (ICASSP 2022).
- Kun Wang
Conservative Contextual Combinatorial Cascading Bandit (IEEE Access Volume 9, 2021).