I’m a third-year Ph.D. student at Boston University Center for Computing and Data Sciences where I’m fortunate to be advised by Aldo Pacchiano. Previously, I did my master in the Robotics Department at the University of Michigan. I also spent time in Honda Research Institute US as a Research Intern in Summer 2022

I am interested in developing multi-task and continual-learning agents. To this end, I study reinforcement learning and (nonstationary) bandits in both online and offline settings. Recently, I have been particularly excited about representation learning in RL, as I believe that learning effective representations is key to building such agents.

Manuscripts Under Review

Publications

• Yichen Song, Alessio Russo, Aldo Pacchiano, “Future Information-Directed Sampling for Bayesian Nonstationary Bandits”, ICML 2026 Workshop on Decision-Making from Offline Datasets to Online Adaptation: Black-Box Optimization to Reinforcement Learning.
• Alessio Russo, Yichen Song, Aldo Pacchiano, Pure exploration with feedback graphs, International Conference on Artificial Intelligence and Statistics (AISTATS) 2025 (Oral presentation).
• Hongyu Zhou, Yichen Song, Vasileios Tzoumas, Safe non-stochastic control of control-affine systems: An online convex optimization approach, IEEE Robotics and Automation Letters (RA-L), 2023.