What is it like to see? Models of visual awareness in brains, experiments and machines

The scientific study of consciousness represents one of the most intriguing challenges in modern science. Over the last three decades, the methodological advancement in neuroscience has led to substantial improvement in the understanding of how the human brain builds conscious repre- sentations of the outer and inner world, going beyond metaphysical and philosophical arguments about the nature of consciousness. However, we’re still far from a comprehensive description about how consciousness emerges. A plethora of theories of consciousness (Toc) is trying to link neural correlates of consciousness-related phenomena to mechanisms which lead to the latter, but a robust link between a theoretical formalization of the mechanisms underlying conscious experience and its neural basis is still missing. In this context, model-based approaches to the scientific study of consciousness represents an optimal approach to explain well-defined sub- problems of consciousness, providing mechanistic explanations of experimental observations in terms of model of consciousness (MoC). In this thesis I will try to describe a possible method- ological roadmap to assess the study of consciousness, isolating specific problems and assessing them toward the synergistic integration of different modeling approaches. At first I will try to define a precise taxonomy of the type of models that can be exploited in neuroscience, namely neuropsychological and explicit and implicit computational models, providing a description of their integration with experimental paradigms. Then, I will focus on the specific problem of vi- sual awareness, an optimal subdomain for a mechanistic investigation of consciousness. Through the presentation of the four main studies conducted during my PhD I will outline how combining computational modeling, neuropsychology, and targeted experimental manipulations can refine existing MoC and generate empirically grounded insights into the neural mechanisms underlying conscious visual perception. Together, these studies exemplify how model-based approaches, in- formed by appropriate theoretical frameworks, can advance our understanding of consciousness by yielding increasingly precise and testable descriptions of specific conscious-related phenomena.