Ricerc@Sapienza

My research aims to understand how the dynamic organization of distributed neural networks supports flexible, high-level cognitive functions such as voluntary motor control, selective attention, and memory-guided decision-making. To address this challenge, I adopt an integrated systems neuroscience perspective that combines experimental neurophysiology, theoretical modeling, and analytical frameworks inspired by statistical physics, graph theory, machine learning and information theory.

A central focus of my work is the study of multi-scale cortical dynamics during behavioral paradigms that isolate core cognitive processes. These include response inhibition and memory-based reasoning. Using high-resolution, multi-site electrophysiological recordings in non-human primates and human subjects performing these tasks, I investigate how neural populations interact over time to implement flexible control strategies and generalizable decision rules.

In parallel, I develop computational models and analytical tools to characterize the spatiotemporal organization of neural activity. These models—rooted in statistical mechanics and complex network theory—aim to reconstruct the functional architecture of task-relevant circuits and identify interpretable signatures of cognitive operations. This approach aims to bridge data-driven analysis and theoretical formalism, contributing both to the understanding of brain function and to the advancement of neurotechnologies such as brain-computer interfaces (BCIs).

Over the past years, I have extended this framework toward explainable artificial intelligence (XAI) and to the design of biologically inspired neural architectures that integrate predictive coding, symbolic reasoning, and self-supervised learning. The goal of these models is not only to decode brain activity but also tryng to emulate and and interpret the underlying computational strategies, providing a common ground between neuroscience and interpretable AI.

Most recently, my work is expanding toward cross-species comparisons. In collaboration with national and international clinical partners, I am collecting stereo-EEG data in human subjects during cognitive tasks analogous to those used in primates. These recordings serve the dual purpose of investigating conserved cognitive mechanisms and supporting clinical efforts to localize seizure onset and propagation networks during pre-surgical monitoring for drug-resistant epilepsy.

In summary, my research goal is to contribute to a unified theoretical and methodological framework linking neural dynamics, cognitive function, and machine learning—advancing our understanding of how the brain flexibly encodes, processes, and acts upon information in complex environments.