This event will be held virtually. Please click here to join. 

The Computational Structure of Human Social Cognition

Abstract: Social cognition is implicated in virtually all areas of uniquely human intelligence, and characterizing its computational structure is a central challenge in cognitive science. In recent decades, social cognition has been increasingly conceptualized as a problem of belief and desire attribution grounded in recursive social inference. I will argue that this view captures only a narrow dimension of the socio-cognitive toolkit. Instead, I will propose that much of human social intelligence is supported by mental models of others as computational engines—probabilistic programs that emphasize representing and tracking real-time computations happening in other people’s minds as they interact with the world. I will present a computational framework of this approach and show how it helps explain uniquely human social behavior.

Speaker Bio: Julian Jara-Ettinger is an assistant professor of psychology at Yale University, with affiliations to the Computer Science department, the Cognitive Science program, and the Wu Tsai institute. Julian received his bachelor’s degree in physics and mathematics at the Universidad Michoacana in Mexico and his PhD in Cognitive Science at MIT. At Yale, Julian’s research group—the computational social cognition lab—aims to characterize the representations and computations that support human social cognition, understand how they emerge and develop, and use them to build more human-like machine social intelligence.