AI · MACHINE LEARNING · ROBOTICS
Collaborative AI for Machines & People
A research group at Purdue University building collaborative AI and robots that continue learning long after we stop teaching them.
Research
Agents that never stop learning.
Introspective agents capable of self-improvement.
Interpretability & introspection
Agents that understand their own knowledge and capabilities, and use it to learn more effectively.
Sample-efficient learning
Learning more with less data through better decision making and model updates.
Theory of mind & personalization
Inferring the beliefs, intentions, and goals of people and other agents.
Featured projects
Recent news
📄 Energy-Based Transfer for RL accepted to CoLLAs 2026.
📄 Concept-Driven Exploration accepted to IROS 2026.
📄 HyperAdapt accepted to TMLR.
📄 GRAIL accepted to ACL 2026.
🏆 HyperAdapt wins an outstanding paper award at the NeurIPS CFFM workshop!
🏆 Joseph receives an Amazon Research Award.
Muhan Lin, Fiona Xie, and Ayush Karmacharya join the lab as new PhD students. Welcome!
📄 Model-Agnostic Policy Explanations accepted to COLM 2025.
🏆 Let Me Help You! wins best paper award at the ICRA Non-Verbal Cues workshop!