Samuel J. Gershman

Papers

Total Citations

123

H-Index

About

Samuel J. Gershman is a leading computational cognitive scientist whose research bridges reinforcement learning, Bayesian inference, and cognitive neuroscience to uncover the fundamental algorithms underlying human intelligence. His most influential work centers on how the brain discovers hierarchical representations for efficient planning, demonstrating that humans spontaneously organize environments into clusters of states, enabling complex problem-solving through abstraction and decomposition. This landmark 2020 paper, with over 100 citations, has reshaped our understanding of how people tackle challenging tasks by breaking them down into manageable sub-problems across multiple levels. Gershman's broader contributions include developing normative models of learning, memory, and decision-making that integrate Bayesian principles with neural mechanisms. His work has profound implications for artificial intelligence, particularly in designing agents capable of human-like hierarchical reasoning. As a professor at Harvard University, Gershman continues to pioneer the intersection of cognitive science and machine learning, earning recognition as a rising star in computational neuroscience. His research not only explains how humans navigate complex environments but also provides a blueprint for building more flexible, efficient AI systems.