Jacob Beck is a research scientist in the Machine Learning Research Group with a background in reinforcement learning. He's contributed significantly to meta-reinforcement learning, including applications to hypernetworks, in-context learning, and protein fitness prediction – in addition to authoring a survey of meta-RL. He’s also worked on human feedback, imitation learning, and multi-agent reinforcement learning.

Currently, he is interested in agents for coding and creating agentic foundation models that solve a diverse set of tasks (i.e., broad generalization) while also learning efficiently from limited experience (i.e., rapid adaptation). Previously, he did his PhD at the University of Oxford, his MS and BS at Brown University, and a pre-doc at Microsoft Research.

Personal Website: https://www.jakebeck.com