Chelsea Finn is an assistant professor of computer science and electrical engineering at Stanford University, where she leads the IRIS lab studying intelligence through robotic interaction at scale. She holds a B.S. in electrical engineering and computer science from MIT and a Ph.D. from UC Berkeley, and previously worked at Google Brain.
Finn is best known for Model-Agnostic Meta-Learning (MAML), introduced in 2017 with Pieter Abbeel and Sergey Levine. MAML trains a model so that a few gradient steps on a small amount of data from a new task produce good performance, a clean and widely used formulation of “learning to learn.” The work won the ACM Doctoral Dissertation Award and helped make meta-learning a mainstream research area, with applications across robotics, few-shot classification, and reinforcement learning. She later co-authored Direct Preference Optimization (DPO), a simpler alternative to RLHF for aligning language models.
In 2024 Finn co-founded Physical Intelligence, a startup developing general-purpose foundation models for robots, alongside collaborators including Sergey Levine.