The large-scale organization of object representation

Talia Konkle

Talia Konkle

(Harvard University )

Play Video

Duration: 1h 15m 6s

Date: January 14, 2015


How are representational spaces of the mind mapped onto the surface of the brain? In the domain of visual object representation, there is no widely accepted model of the coding dimensions of objects: the representational space of objects is so high-dimensional that its neural organization is assumed to have only a few category-selective regions with remaining heterogenous structure only at a fine-grained scale.

In contrast to this view, I will present results showing there is a large-scale organization of object responses spanning the entire ventral and lateral occipito-temporal cortex, based on two major cognitive dimensions of objects: animacy and real-world size. Regions selective for faces, bodies, and scenes fit within this organization at a meso-scale. These results demonstrate that object cortex, just like early visual cortex, has structure that can be explained at multiple spatial scales, where the large-scale neural organization reflects core factors in the representational space.

Motivated by these findings, I will present a theoretical framework in which we can leverage the spatial topography of neural responses to help infer and test the structure of the underlying cognitive architecture. While developed in the context of explaining object representation, this framework is broadly applicable to understanding the organization of many forms of knowledge on the cortical surface.

Created: Wednesday, January 14th, 2015