Nitish Dashora

Nitish Dashora is a PhD student in EECS at MIT, where he researches robot learning, continual learning, and foundation models as a member of the Improbable AI Lab at CSAIL under the supervision of Professor Pulkit Agrawal. He previously conducted research at UC Berkeley's Robotic AI and Learning Lab (BAIR) under Professor Sergey Levine, where he worked on deep reinforcement learning, unsupervised pre-training, and autonomous navigation. His published work includes ViNT, a foundation model for visual navigation, and ViVa, which proposes video-trained value functions for guiding online reinforcement learning.