Séminaire ayant lieu dans le cadre du doctorat en informatique cognitive, en collaboration avec le centre de recherche CRIA    

TITRE :  Human vs Machine in the Game of Hidden Rules

Jacob FELDMAN

Jeudi le 5 mars 2026 à 10h30

Local PK-5115 (Il est possible d'y assister en virtuel en vous inscrivant ici)               

RÉSUMÉ

Comparisons of human and machine intelligence are often grounded in supposition, unencumbered by empirical data about human performance. In this talk I'll present results comparing human and machine performance in on a common platform, the "Game of Hidden Rules" (GOHR). The GOHR is a simple rule-discovery game in which a player---human or AI---tries to classify objects into categories based on an unknown rule that they must infer by trial and error. Human players solve such problems about two orders of magnitude faster than (blank slate) AI models. In general, human and AI performance are almost completely uncorrelated, suggesting that contemporary AI does not yet effectively reflect the way that humans learn.

BIOGRAPHIE

Jacob FELDMAN is Professor of Psychology and Cognitive Science at Rutgers University, where he directs the Visual Cognition Lab. His research focuses on computational models of human visual perception and concept learning, particularly perceptual organization, shape representation, and categorization. Feldman has worked on the simplicity principle in human concept learning and Boolean complexity minimization, as well as on Bayesian models of perception and learning.

RÉFÉRENCES 

Feldman, J. (2025). Simplicity and complexity of probabilistically-defined concepts. Psychological Review, in press. 

Feldman, J. (2024). Probabilistic origins of compositional mental representations. Psychological Review, 131(3), 599-624. 

Destler, N., Singh, M., & Feldman, J. (2023). Skeleton-based shape similarity. Psychological Review, 130(6), 1653-1671. 

Feldman, J. (2021). Information-theoretic signal detection theory. Psychological Review, 128(5), 976-987.