Stefano Panzeri

This study shows that endogenous oxytocin alters brain network dynamics, increasing salience-related network states and decreasing sensory-dominated states in mice.

The paper shows humans use others' movement to infer confidence and guide collective decisions, improving accuracy when confidence is intermediate.

The paper uses PID to analyze how stimulus information is distributed and shared across cortical networks in ferrets during perceptual decisions, finding that faster decisions involve more redundant information across functionally coupled sites.

The paper introduces MICs to analyze how modulators reshape interactions between neural populations using a bilinear perturbation framework, showing its effectiveness in simulations and real neural data.

The paper shows that cortical excitability inversely affects fMRI connectivity, with low-frequency neuronal coupling being the key driver of these changes.

The paper develops a whole-cortex model of resting-state fMRI activity in mice, showing that attractor dynamics explain the emergence and structure of co-activation patterns (CAPs) through anatomical connectivity.

The paper investigates whether cortical neurons projecting to the same target form specialized population codes, finding structured correlations that enhance information transmission during correct behavioral choices.

This study shows that information gain in goal-directed learning is encoded through synergistic and higher-order functional interactions across cortical circuits.

The paper shows that brain activity in the parietal cortex encodes evidence consistently with prior choices, leading to confirmation bias in decision-making.

This paper corrects optical aberrations in GRIN lens-based microendoscopes to improve imaging performance in deep brain regions.

MINT is a toolbox for analyzing multivariate neural information coding and transmission using information theory and dimensionality reduction, applied to simulated and real neural data.

This study shows that efficient coding principles can explain key properties of excitatory-inhibitory neural networks, including their connectivity and balance of excitation and inhibition.

The paper shows that cortical astrocytes mediate learned behavior by extending norepinephrine signals through calcium dynamics and ATP release, influencing neuronal encoding of task variables.

The paper identifies evolutionarily conserved dynamic fMRI network patterns across mouse, macaque, and human brains using frame-wise clustering of time-series data.

The paper models how spiking neural networks efficiently process dynamic sensory features across layers, showing that feature mixing in neurons reduces metabolic cost and matches empirical observations in the mouse somatosensory system.

The paper develops bias correction methods for Partial Information Decomposition (PID) in neural data, improving estimates of synergy and redundancy in mouse brain regions.

This study shows that sensory information increases in primary somatosensory cortex during learning, while choice information emerges later and relies on population coding.

The paper investigates how synaptic connectivity in the posterior parietal cortex supports decision-making by identifying opponent inhibition motifs between excitatory and inhibitory neurons.

ASTRA is a deep learning tool for automated segmentation of astrocytic calcium signals in two-photon imaging, outperforming existing methods and enabling large-scale analysis.

The paper identifies neurons in V1, PPC, and RSC that encode mixed selectivity for navigation decisions, forming efficient population codes in RSC for correct choices.

The paper measures redundant and synergistic functional connectivity in mouse auditory cortex during a perceptual task, finding that connected neuron pairs show more redundancy and less synergy than unconnected pairs.

This paper introduces a new information-theoretic measure, FIT, to quantify feature-specific information transfer between brain regions using simulated and empirical neural data.

NIT is an open-source toolbox for information-theoretic analysis of neural data, including calcium imaging, offering tools for encoding quantification, information transmission, and partial information decompositions.

This study investigates population coding strategies in human tactile afferents, finding that optimal density and composition depend on tactile features, and temporal information is more critical than spatial information.

The paper develops a method to classify network states from local field potentials in awake mice, showing that population signals predict single-cell activity across different network states.

The paper investigates how neuronal networks in mouse auditory cortex encode behavioral choice during pure-tone discrimination tasks, finding that specialized neurons transmit task-relevant information sequentially.

This study shows that astrocytic calcium signals in the hippocampus encode spatial information complementary to neurons, enhancing spatial position decoding.

The paper shows that silencing the prefrontal cortex paradoxically increases resting-state fMRI connectivity with its thalamo-cortical terminals.

This study investigates spatiotemporal dynamics of spontaneous brain activity in awake mice using resting-state fMRI, finding unique patterns distinct from anesthetized states.

The paper shows that neural correlations in association cortex enhance behavioral performance by improving information consistency, despite limiting sensory information.