About Me

Hello! I'm Calvin Luo, a researcher based in Stanford, CA. I recently graduated my PhD from Brown University, where I was advised by Chen Sun. I spent the final year of my PhD visiting the Robotics and Embodied AI Lab at Stanford University, advised by Shuran Song. I also work closely with Yilun Du.

In the past, I was an AI Resident at Google in Mountain View, where I worked with Ting Chen, Hossein Mobahi, and Samy Bengio. Before that, I was an undergraduate at the University of Toronto, where I was advised by David Duvenaud.

Research

I study how intelligent robotic decision-making can be unlocked, even without large-scale fully-labeled robotic datasets. In particular, my research presents internet-pretrained visual generative models as powerful and flexible sources of decision-making capabilities for robotics. My focus is on two aspects of robotic intelligence:

• Zero-Shot Generalization: How can a robot increase its initial success rate for novel tasks and settings?
• Sample-Efficient Self-Improvement: How can a robot decrease the amount of attempts it needs to refine its behavior?

My work demonstrates how visual generative models can offer significant generalization and self-improvement benefits for robotics, and covers themes such as adaptation, exploration, and inter-model collaboration. Beyond robotics, I envision a future where models can collaborate effectively with each other, bridging modality and information gaps to unlock greater intelligent capabilities as a collective.

News!

• (04/26) One paper accepted to ICML 2026 - see you all in Seoul!
• (01/26) One paper accepted to ICLR 2026 - see you all in Rio de Janeiro!
• (09/25) One paper accepted to NeurIPS 2025 - see you all in San Diego!
• (06/25) I received the Research Mobility Fellowship to visit Stanford via IvyPlus - thank you Brown!
• (01/25) One paper accepted to ICLR 2025 - see you all in Singapore!
• (09/24) One paper accepted to NeurIPS 2024 - see you all in Vancouver!
• (09/24) I am delighted to serve as co-instructor of CSCI2470: Deep Learning for Fall'24.

Selected Work