Python, NumPy, Matplotlib, SciPy, seaborn
Integrated RL algorithms with latent cause inference for continual learning. Work done with postdoc Sashank Pisupati and Prof. Yael Niv at Princeton.
Simulated a grid world with obstacles and rewards, navigated by an actor-critic agent (both value and policy representations). The agent learns various latent states, allowing it to segregate learning for continual learning. Our work has applications in psychiatry — understanding failures in human learning.
I presented my research via oral and poster presentations at the Leadership Alliance National Symposium, 2022.