C02 Riichiro Hira

C02　Dynamics of multiple predictions by cortico-cortical communication

Using the large field-of-view two-photon microscope Diesel2p (Yu et al., 2021), our group is able to measure neural activity in a large region of the dorsal cortex of head-fixed mice. This field-of-view includes ACC, M2, M1, S1, PPC, RSC, and V1. It provides information about not only the activity of neuronal populations in these areas but also the interactions between them. The hierarchical structure among cortical areas, especially in the visual cortices, has conventionally attracted particular attention. It is well-known that this led to applications in deep learning. On the other hand, the cortical regions that we will observe cannot be uniquely described by hierarchical relationships. It is thought that a variety of multimodal information is context-dependently transferred between the areas. The purpose of this study is to describe such interactions between cortical areas as complex and multiple predictive relationships. In particular, we will elucidate the mechanism by observing and controlling how different types of information, such as sensory, location, and value information, multiply predict each other through interactions between the cortical areas.