Information dynamics in neuronal networks of macaque cerebral cortex reflect cognitive state and behavior

The brain is often described as “processing” information: somehow, the decentralized interactions of billions of neurons collectively are somehow able to give rise to “emergent” behaviors, such as perception, cognition, and action. In neuroscience and cognitive science, however, “information processing” is often vaguely defined, making an exact model connecting neurodynamics, information processing, and behavior difficult to pin down. While considerable previous work has examined the structure of information dynamics in cultures or models, it remains uncertain what insights information dynamics can provide about cognition and behavior in order to interact with the environment. In this paper, we use information theory and the theory of information dynamics as a formal framework to explore information processing in multi area neuronal networks recorded from three macaques engaged in sensory-motor transformations: perceiving a visual cue, preparation of a grasping movement, and movement execution. We found that different states and grasp conditions are associated with significant re-configurations of the effective network structure and the overall information flowing through the system. Crucially, differences between cognitive and behavioral states and conditions were related to changes to higher-order, synergistic information dynamics not localizable to a single pair of source/target neurons. Our results suggest that the combined application of information-theoretic analysis of dynamics and network science inference to networks of neurons is a powerful tool to probe the neuronal basis of cognition and behavior. ### Competing Interest Statement The authors have declared no competing interest.