Abstract
Mathematics is among humanity's most remarkable achievements, yet we still lack a comprehensive understanding of how the brain performs even simple arithmetic. In this talk, I will present a series of studies investigating the encoding of elementary math, as well as the architecture, spatiotemporal dynamics, and causal role of the underlying brain networks. I will show that arithmetic computations selectively activate a distinct network in the human brain, which dissociates from language areas and overlaps with regions related to object recognition, visuospatial attention, working memory and relational reasoning. Next, using machine learning and intracranial recordings in humans, I will demonstrate how we can precisely track the cascade of unfolding representational codes during mental arithmetic, shedding light on the roles of each hub of the math network. Overall, this talk will provide insights into how elementary math concepts are implemented in the brain and, more broadly, show how the case study of math cognition can help us understand the algorithms of human intelligence.
