I am a PhD student at the Autonomous Learning Lab at the University of Massachusetts, Amherst. I am advised by Bruno Castro da Silva and Philip Thomas. I am interested in applying reinforcement learning to practical applications, and my research focuses on addressing the challenges associated with this. To that end, some representative problems that interest me include robust policy evaluation, abstractions in complex sequential processes, studying the effects of partial observability, and exploring analogies between large language models and reinforcement learning.
Previously, I was a master’s student at Carnegie Mellon University in the Department of Electrical and Computer Engineering. During my time there, I worked on reinforcement learning with Prof. Ben Eysenbach and bandit algorithms with Prof. Gauri Joshi. I recieved my Bachelor’s degree in Electrical Engineering from IIT Madras, where I worked on reinforcement learning with Prof. L.A. Prashanth.
In my free time, I enjoy climbing, trail running and playing as well as watching football (soccer).
Updates
- October 2024: Paper accepted at NeurIPS-24! STAR is a new framework for OPE that leverages state abstraction.
- May 2024: Started as a Research Scientist Intern at Waymo.
- April 2024: Our analysis and survey on reinforcement learning from human feedback (RLHF) is up on arXiv.
- December 2023: Two papers accepted at AAAI-24! Distributional OPE for Slate Recommendations and Rethinking Eligibility Traces.
- November 2023: Gave a talk on off-policy evaluation under partial observability at A*STAR's Center for Frontier AI Research (CFAR).
- October 2023: Our work on model-based off-policy evaluation under partial observability was accepted at the RealML workshop at NeurIPS' 23.
- August 2023: Our work on distributional off-policy evaluation for slate recommender systems is up on arXiv.
- August 2022: Awarded the Robbin Popplestone Graduate Fellowship.
- May 2022: Started as a Research Scientist Intern at Adobe Research.