Dr. Lusha Zhu, Peking University

Date/Time

Friday, December 8, 2023, 16:30-18:00 (JST) (15:30-17:00 CST)

Abstract

Inferences in social interactions: from social learning to knowledge alignment

In a rock-paper-scissor game, how does an individual predict which action her opponent will select? This inferential ability is probably one of the most complex functions of human intelligence, and its pathological expressions are believed to be closely related to social abnormalities seen in a range of psychiatric conditions.

In this talk, I will discuss three computational mechanisms likely underlying the inferential ability –– by learning from direct experience, simulating others based on rational decision-making models and, somewhat surprisingly, aligning knowledge representations that helps avoid social reasoning. I will present examples from our past and ongoing research in which we combined game-theoretic paradigms, computational modeling, and functional neuroimaging to explore the neurocomputation underlying these processes. I will discuss how the integration of these processes may facilitate flexible inferences about social partners, even in novel interactions that a decision-maker has never before directly experienced.

For inquiries

cbs-academic[at]ml.riken.jp