Grace Lam

Papers

Total Citations

H-Index

About

Grace Lam is a leading researcher at the intersection of robotics, computer vision, and natural language processing, with a primary focus on developing generalist robot policies. Her most impactful work to date is "OpenVLA: An Open-Source Vision-Language-Action Model" (2024, 39 citations), which introduces a groundbreaking approach to robot learning. Rather than training new behaviors from scratch, Lam's OpenVLA model leverages large-scale, Internet-pretrained vision-language models and fine-tunes them on diverse robot demonstration data. This paradigm shift enables robots to acquire robust, generalizable skills through simple fine-tuning, dramatically reducing the time and data required for new tasks. Her work has already garnered significant attention in the robotics community, with 39 citations in under a year, and promises to democratize access to state-of-the-art robot learning. By open-sourcing the model, Lam has enabled researchers worldwide to build upon her foundation, accelerating progress toward truly versatile, intelligent robots. Her contributions are paving the way for a future where robots can learn from human language and visual examples, rather than requiring specialized programming for every new action.