Max Planck Institute for Dynamics and Self-Organization -- Department for Nonlinear Dynamics and Network Dynamics Group
Personal tools
Log in

Agostina Palmigiano

(Ph.D. Student)
Group: Theoretical Neurophysics
Projects: Phase space structure and chaos of pulse-coupled network dynamical systems, Modelling the role of neural oscillations in information routing
Address:  Am Faßberg 17
37077 Göttingen
Office: Room 1.03
Phone: +49-(0)551-5176-421
Email: send email
Interests: Oscillatory Networks
Delayed dynamical systems
Information Theory
Flexible information routing with transient synchrony
Perception and cognition rely on context-dependent selection of relevant inputs and flexible inter-areal brain communication. However, despite its fundamental role, the basis of dynamic information routing remains an unsolved problem. In particular, the circuit mechanisms that underlie the fast reconfiguration of selective information transfer on behavioral time-scales are poorly understood. We work with a class of network models in which neurons spike irregularly while the collective activity exhibits an irregular rhythm in the gamma band. This regime, in which short transient episodes of higher synchronization occur spontaneously, reproduces the frequency and power fluctuations found in cortical measurements. In order to determine whether this variability and irregularity presents an advantage or a limitation to flexible information transmission, we profit form small motifs of two or three interconnected areas. When several circuits with these characteristics are connected by long range excitatory connections, the activity of the areas coordinate to spontaneously give rise to simultaneous gamma bursts of sparsely synchronized activity. Within each gamma burst, the areas transiently phase-lock, reproducing in the short bursts the out of phase locking patterns observed in higher synchrony regimes. Profiting from state resolved information theoretical tools we found that these transient patterns of coherence gate the information flow between the areas. Information flows in a direction determined by the phase relation and with a strength regulated by the fluctuating level of coherence. Externally injected inputs, representing sensory information or feedback from surrounding areas, are transmitted following the paths found by the analysis of the spontaneous activity. Weak external biases can modulate the pattern of relative phases, favoring particular directions of information transmission.
High dimensional chaos and synchronous irregular dynamics in delayed spiking networks


The dynamical stability of neuronal networks, and the possibility of chaotic dynamics in the brain pose profound questions to the mechanisms underlying perception. Whether the rapid amplification of differences of nearby states and the finite-time dissolution of encoded information offer an advantage for brain processing or works in detriment to it, remains to be understood. We advance on the tractability of large, highly heterogeneous neuronal networks of exactly solvable neuronal models with delayed pulse-coupled interactions and study its stability properties. We develop a framework in which pulse-coupled delayed systems with an infinite dimensional phase space can be studied in an equivalent system of fixed and finite degrees of freedom. By introducing a single compartment axon for each neuron, delays arise from the extra steps of integration needed for the variable to reach threshold. The Jacobian of the equivalent system can be analytically obtained and numerically evaluated in event-based simulations, giving access to an exact calculation of the entropy production rate of the delayed network. Profiting from neuronal models with a tunable action potential onset rapidness, find that depending on the speed of the action potential onset and the level of heterogeneities, the asynchronous irregular regime characteristic of balanced state networks  loses stability with increasing delays to either a slow synchronous irregular or a fast synchronous irregular state for slow and fast onset rapidness respectively. In networks of neurons with slow action potential onset, the transition to collective oscillations leads to an increase of the exponential rate of divergence of nearby trajectories and of the entropy production rate of the chaotic dynamics, revealing that chaos is intensified in the slow synchronous irregular state. The attractor dimension, instead of increasing linearly with increasing delay as reported in many other studies, decreases until eventually the network reaches full synchrony.

Publications (chronological - click here to see all publications)