Christopher Sims

Chris Sims received a B.S. in computer science from Cornell University (2003) and a Ph.D. in Cognitive Science from RPI (2009). He then held a postdoctoral research position at the University of Rochester and a faculty position at Drexel University before rejoining RPI in 2017.

Sims directs the Adaptive Computational Cognition (AdaCog) Laboratory, where his research develops computational models of human cognition, with a focus on learning, decision making, and the computational principles that connect human and machine intelligence. A central contribution of his work is the application of information theory to explain why cognitive systems are structured the way they are. In a 2018 study published in Science, Sims showed that the “universal law of generalization” — which describes how organisms generalize from past experiences — can be derived from principles of efficient information coding, suggesting that human perceptual systems are tuned to make the best use of limited information-processing capacity. He has extended this framework in subsequent work, including a 2025 study in Nature Communications showing that people learn generalizable representations through efficient coding, and a 2024 paper in Psychological Review examining how efficient visual representations support learning and decision making.

A primary focus of Sims's research is the intersection of computational cognitive science and artificial intelligence, treating human and machine learning as a virtuous circle in which each field informs the other. His work examines how cognitive and computational constraints shape intelligent behavior, and what this reveals about the gap between current AI systems and human intelligence. This includes work on computationally rational models of reinforcement learning, as well as the broader question of “missing ingredients” that separate today's AI from the flexibility and efficiency of human cognition.

His research has appeared in journals including Science, Nature Communications, Psychological Review, Scientific Reports, and has been cited over 2,000 times.