Subramanian's primary research interest is in multi-agent systems. He is interested in the issues of scale, non-stationarity, effective communication, safety, and sample inefficiency in multi-agent learning systems. As a consequence, his area of work is at the intersection of Reinforcement Learni...
Subramanian's primary research interest is in multi-agent systems. He is interested in the issues of scale, non-stationarity, effective communication, safety, and sample inefficiency in multi-agent learning systems. As a consequence, his area of work is at the intersection of Reinforcement Learning and Game Theory. He is also highly interested in the theoretical aspects of Reinforcement Learning. Subramanian has two important and related long-term research goals. The first is to bridge the widening gap between the theoretical understanding and empirical advances of multi-agent reinforcement learning. The second is to make multi-agent learning algorithms applicable to a variety of large-scale real-world problems.
He is particularly interested in research applications of reinforcement learning and multi-agent reinforcement learning in large language models (generative AI), robotics, finance, recommender systems, and autonomous driving.
Subramanian is also affiliated with the Vector Institute for Artificial Intelligence, Toronto, and the Schwartz Reisman Institute for Technology and Society, Toronto. He serves as a mentor at the Indigenous Black Engineering and Technology (IBET) PhD Project.