Dynamic networks

We study the brain as a network whose effective connections change with sensory input, behavioral state, and the internal dynamics already in motion. Our work shows that stimulus type reshapes functional-network topology and that the balance among stimulus drive, behavior, and internal dynamics changes across brain states and the visual hierarchy. In our new bioRxiv preprint, Structured Sparsification of Signal-Transmission Networks Enhances Visual Information Coding, we find that locomotion is associated with fewer, more local, modular, feature-specific, and feedforward interactions, together with faster and more faithful sensory coding; rate-based models link these improvements to reduced shared variability and more feedforward signal transmission. We now ask which functional connections remain stable across states, how arousal and behavior select different communication pathways, and how feedforward, recurrent, and cross-area interactions balance coding speed with robustness.