Boyuan Chen is a PhD alumnus at MIT EECS, known for his work in AI and robotics. Now working as a scientist at OpenAI, Boyuan is one of the five researchers training GPT image generation and a member of the Sora video generation team. He earned his PhD in Electrical Engineering and Computer Science from MIT, with a minor in philosophy. His research focuses on world models, embodied AI, and reinforcement learning—integrating these directions to build AI systems that can better understand and act in the physical world.

Chen has authored several influential papers recognized in both academia and industry, including Diffusion Forcing, SpatialVLM, and History Guidance. His work has been featured at top machine learning and robotics conferences as well as in media outlets. He was a recipient of the Seneff-Zue CS Fellowship and has previously been affiliated with Google DeepMind, where he worked on foundation models for decision making. At MIT, he was a member of the Scene Representation Group led by Professor Vincent Sitzmann and the  Robot Locomotion Group led by Professor Russ Tedrake.

Before joining MIT, Boyuan obtained his bachelor’s degree in computer science and math at UC Berkeley, where he spent a signficant amount of time doing research at Berkeley Artificial Intelligence Research (BAIR) on deep reinforcement learning and unsupervised learning. He also spent a year studying philosophy during my undergrad. Boyuan is a big fan of chess, robots and boba.

For more information, please visit the following links:

• Boyuan Chen Homepage
• Twitter
• Github
• LinkedIn

Boyuan's academic profiles:

• Google Scholar
• OpenReview
• DBLP
• ORCID