Hi there! I’m Gokul, a PhD student in the Robotics Institute at Carnegie Mellon University working on learning algorithms for interactive robots.
I work with Steven Wu and Drew Bagnell. I completed an M.S. at UC Berkeley under Anca Dragan and Sergey Levine. My thesis was focused on Learning with Humans in the Loop.
I’ve spent summers working as a SWE Intern @ Intuit, Data Engineering Intern @ SpaceX, Autonomous Vehicles Perception Intern @ NVIDIA, and Motion Planning ML Intern @ Aurora.
In my free time, I do origami, hackathons, and run. I’m a huge fan of birds (especially lovebirds), books (especially those by Murakami), and indie or alt. music (especially that of Radiohead).
Events & News
May 2021 - Our paper on a unifying framework, efficient reductions, and practical algorithms for imitation learning was accepted to ICML ‘21! You can check our video explanations and code here.
September 2020 - I started a PhD at CMU’s Robotics Institute!
May 2020 - I finished up my M.S. at UC Berkeley, with more than a little help from my wonderful friends and collaborators!
Research Highlights
Of Moments and Matching: A Game-Theoretic Framework for Closing the Imitation Gap
We construct a taxonomy for imitation learning algorithms, derive bounds for each class, construct novel reduction-based algorithmic templates that achieve these bounds, and implement simple elegant realizations with competitive emperical performance. Published at ICML 2021. [Website]
Scaled Autonomy
As fleet sizes grow, it becomes difficult for a single teleoperator to supervise all robots. We learn from user preferences to make automated switches. Our work was published at ICRA 2020.
On the Utility of Model Learning in HRI
We compare three methods of modeling human driving behavior on their sample efficiency and transferability. Our work was published at HRI 2019.