Understanding how the brain combines external sensory information with internally generated signals to produce conscious experiences is a fundamental question in sensory neuroscience. The conventional model proposes that sensory neurons receive external signals through receptive fields, encoding specific aspects of the signal, followed by stochastic spike generation to account for in vivo variability. In this presentation, I will discuss my research focusing on these two key aspects of brain processing.

First, I will present my data-driven modeling approach, which employs naturalistic stimuli and behavioral conditions to refine brain models and uncover encoding properties. I will specifically explore two applications in marmoset monkeys: the retinotopic organization of third-tier visual areas using widefield optical imaging, and multimodal integration in the posterior-parietal cortex using chronic electrophysiology.

In the latter part of my talk, I will briefly discuss my previous study on the apparent stochasticity in neuronal responses. Contrary to the prevailing notion that ongoing firing underlying response stochasticity is purely random, my work with widefield voltage and calcium imaging in mice has revealed spatiotemporal order in ongoing activity, contingent upon the animal's brain state.

By investigating both the processing of external information and internally generated signals, we can gain valuable insights into the input-output relationship of the brain, ultimately advancing our understanding of conscious experience.