Ben Eysenbach, an expert in reinforcement learning, has won a National Science Foundation CAREER Award, one of the top honors for early career faculty.
The award is part of NSF’s Faculty Early Career Development Program and supports junior faculty who exemplify leadership in education and research. It comes with nearly $600,000 in research funding over five years.
Eysenbach, an assistant professor of computer science, will use the grant to study how AI models can learn complex new tasks. Currently, models are trained using reinforcement learning techniques, a type of machine learning that uses rewards and feedback to teach an autonomous agent to make intelligent decisions through trial and error. But complex tasks that require dozens of skills are difficult to teach using this step-by-step approach.
Eysenbach’s work focuses on developing new reinforcement learning techniques that enable AI models to learn through exploration and experimentation. These skills can then be applied across multiple tasks, making the AI models better at learning complex tasks. More efficient algorithms not only train an AI model more quickly, they also require much less computing power.
Eysenbach joined the Princeton faculty in 2023. He is affiliated with the Princeton Program in Cognitive Science, the Princeton Language Initiative and the Natural and Artificial Minds initiative.
Eysenbach’s work has been recognized by a Sloan Fellowship, a NeurIPS Best Paper Award and a Hertz Fellowship. He has also received an Alfred Rheinstein Faculty Award and recognition for outstanding teaching from the School of Engineering and Applied Science. He completed a doctorate in machine learning at Carnegie Mellon University and spent several years at Google Brain and Google Research before and during his doctoral work. He has an undergraduate degree in mathematics from MIT.