Stable Offline Value Function Learning with Bisimulation-based Representations
[arxiv]Brahma S. Pavse, Yudong Chen, Qiaomin Xie, Josiah P. Hanna
I am a third-year Computer Science PhD candidate at the University of Wisconsin-Madison.
I am broadly interested in representation learning and abstractions for reinforcement learning. Poorly learned representations can lead to data-inefficient learning, instability, and high variance. My works studies how we can design representation learning objectives that can enable an RL agent to reliably make predictions about its environment for validation and control.
I am fortunate to be advised by Josiah Hanna, and I also closely collaborate with Qiaomin Xie and Yudong Chen. I have also interned at Sony AI.
Previously, I completed my BS and MS in Computer Science from the University of Texas at Austin, where I was fortunate to be advised by Peter Stone. I also worked as a software engineer at Salesforce and SAS Institute.
Feel free to shoot me an email if you want to chat!