- Paper 'From Foresight to Forethought: VLM-in-the-loop Policy Steering via Latent Alignment' accepted to RSS 2025
- Paper 'Learning Generalizable Tool-use Skills through Trajectory Generation' accepted to IROS 2024
- 'Open X-Embodiment' won Best Paper Award at ICRA 2024
- Two papers 'DROID' and 'HACMan++' accepted to RSS 2024
- Bimanual manipulation work accepted to CoRL 2023 as an Oral presentation
Research Experience
- Ph.D. student at CMU Intent Lab, focusing on open-world robot learning and human-robot interaction
- Previously worked with Prof. David Held on generalizable methods for long-horizon contact-rich manipulation
- Focused on assistive feeding and bimanual manipulation during master's studies at Stanford University
- Collaborated with Prof. Yi Wu from Tsinghua University on reinforcement learning and self-imitation
Education
- Ph.D. student at CMU Robotics Institute, advised by Prof. Andrea Bajcsy
- Master's student at Stanford University, supervised by Prof. Dorsa Sadigh
- Worked with Prof. Yi Wu from Tsinghua University at Shanghai Qi Zhi Institute on reinforcement learning and self-imitation
- Undergraduate research with Prof. Pieter Abbeel and Prof. Lerrel Pinto on reinforcement learning for deformable object manipulation
Background
Research interests include open-world robot learning and human-robot interaction, focusing on overcoming the embodiment gap that limits the application of foundation models to robotics. By grounding high-level semantic reasoning in the continuous dynamic environment of the physical world, aims to build robots that can learn, reason, and act capably in human-centered environments.