About


I am a 5th year Ph.D. student in the Computer Science Department at UT Austin, supervised by Prof. Peter Stone. I am fortunate to be a current student researcher at Google DeepMind in Montreal, and to have previously interned at Sony AI. My PhD work is focused on understanding cooperative dynamics in multi-agent reinforcement learning settings, and on designing algorithms to enable artificial agents to coordinate with previously unseen agents such as humans—a problem known as ad hoc teamwork.

My work in multi-agent reinforcement learning is grounded by a broad set of general research interests. In no particular order, I’m interested in:

  • Understanding the principles of learning for both individuals and groups of agents, especially those necessary for cooperation, and those that result in emergent abilities in rich multi-agent systems.
  • Understanding necessary and sufficient conditions for generalist agents that operate within a rich, multi-agent world, rather than superhuman single-task solvers.

As such, in addition to multi-agent reinforcement learning and ad hoc teamwork, I’m interested in generalization, continual learning, and agents learning from other agents.

I received a B.S. in Mathematics and Computer Science from Duke University, where I researched interpretable machine learning methods for criminal recidivism prediction with Prof. Cynthia Rudin. We also analyzed the COMPAS recidivism prediction algorithm, showing that a nonlinear dependence on age was an alternative explanation for the previously observed racial bias of COMPAS. My work was recognized with a Goldwater Scholarship, a premier national undergraduate research award.