Synchronization transitions in spiking networks with adaptive coupling

Adaptive link sizes is a major breakthrough step in evolving networks and is now considered as an essential process both in biological and artificial neural networks. In adaptive networks the link weights change in time and, in brain dynamics, these changes are controlled by the potential variations of the pre- and post-synaptic neurons. In particular, in biological neural networks the adaptivity of the links (synapses) was first addressed by D. Hebb who proposed the rule that neurons which fire together wire together. In the present study, we explore the effects of adaptive linking in networks where hybrid synchronization patterns (solitaries, chimeras and bump states) are observed in the absence of adaptivity (i.e., for constant coupling strengths). The network consists of Leaky Integrate-and-Fire (LIF) neurons coupled nonlocally in a 1D ring geometry with periodic boundary conditions. The adaptivity follows the Hebbian principle adjusted by the Oja rule to avoid unbounded increase of the coupling strengths. Our results indicate that, for negative coupling strengths, the adaptive LIF network may transit through chimera state regimes with different multiplicity, provided that the time scales governing the link dynamics are relatively small compared to the time scales of the potential evolution. Moreover, the size of the coherent and incoherent domains of the chimera and their multiplicity change following the average coupling strength in the network. Similar results are shown for the case of bump states; for positive coupling strengths, the adaptive LIF network transits through states of different number of active and subthreshold domains. Here also, the size and number of active and subthreshold domains follow the evolution of the average coupling strength. Such transient effects are suppressed when the time scales governing the link adaptive dynamics become of the same order as the time scales governing the potential dynamics. From the point of view of applications, adaptive networks are relevant in brain dynamics and functions such as learning processes, long and short term potentiation, gradual synapses growth in infancy and adolescence, or synapse deterioration due to aging, neurodegenerative disorders and traumas.