Description
Criticality has been proposed as a computationally advantageous operating regime in neuromorphic systems, but its practical implementation is challenged by the need to tune system parameters to a critical point precisely. In disordered systems, structural heterogeneity can generate rare-region effects that modify the nature of phase transitions. Here, we investigate the influence of a bio-inspired hierarchical-modular network topology on the critical dynamics of memristive nanowire networks. To characterise these dynamics, we employ dynamical systems approaches that provide an alternative to conventional statistical methods for identifying criticality. We show that hierarchical-modular networks exhibit distinct transition features absent in conventional self-assembled networks. These results demonstrate that topology strongly influences the collective dynamics and provide a framework for investigating topology-induced rare-region effects in neuromorphic nanowire networks.
| I am the presenting author | Yes |
|---|