Recent theoretical and empirical advances in neuroscience suggest that geometries of neural population codes can implement simultaneous generalization and differentiation. That is, they can allow brains to represent both the general features of categories and, at the same time, maintain a separate representation, as would be needed for category membership. I will explore how these principles can help resolve three outstanding issues in neuroeconomics. First, the need to compute abstract values during evaluation. Second, the need to flexibly bind values with actions during choice. Third, the need to flexibly bind outcomes and the choices that produced them. I will argue that these principles are even more important in naturalistic choices, such as those characterized by continuous and interactive decisions. In particular, I will examine relevance of coding geometry for the context of prey-pursuit problems.

Bio:
Benjamin Y. Hayden is Professor of Neurosurgery (and McNair Scholar) at Baylor College of Medicine. His lab investigates neural mechanisms of reward, decision-making, executive control, and flexibility, often in naturalistic contexts and with human intracranial recordings; he also examines how these neural processes relate to psychiatric disorders (e.g. depression, anxiety, addiction).