Teams at Toyota Research Institute

Conduct research in machine learning for robot manipulation, focusing on trustworthy learning, uncertainty, foundation models, failure detection, and continual learning. The intern will use sensory data and simulation, validate algorithms on real robots, collaborate with researchers and engineers, present findings, and contribute to publications and open-source projects.

Conduct research on post-training and continual learning methods for pretrained generalist robot policies. Develop adaptation techniques using reinforcement learning, imitation learning, policy distillation, world models, and human-in-the-loop learning. Design rigorous simulation and real-robot experiments, analyze results, communicate findings, and contribute to technical reports and publications.

Conduct research on robot learning and human-robot interaction, developing multimodal and hierarchical policies for collaborative robots. Build code prototypes, work with large datasets, run experiments on simulated and physical robots, conduct user studies, maintain robot-data corpora, and contribute to peer-reviewed publications.