Aur, Dorian

This paper develops and characterizes an adaptive hybrid Electrodynamic Intelligence (EDI) Neuro-Patch, a biohybrid interface uniting the physical, electrical, and functional dimensions of neural plasticity. Unlike conventional neuromorphic systems that simulate synaptic learning algorithmically, the hybrid EDI approach treats charge-flux interactions within memristive neurons as the substrate of computation itself. In this framework, axonal conduction and synaptic integration occur from continuous electrodynamic evolution, where memory, adaptation, and signal propagation are inseparable physical processes. The hybrid neuro-patch combines a solid-state conductive upper layer for rapid signal modulation with a lower iontronic layer providing slow, adaptive reconfiguration. Controlled programming pulses induce state transitions, allowing path learning through field-mediated adaptation rather than digital feedback. By integrating devicelevel electrodynamics with cortical network dynamics we establish a quantitative framework linking voltage, conductance, and path reconfiguration across scales from nanometers to cortical circuits. The simulations demonstrate emergent network stabilization and adaptive routing, suggesting that hybrid EDI neuro-patch could provide self-organizing implants for restoring and maintaining stable neural function through the intrinsic physics. By demonstrating that memory, adaptation, and signal propagation are inseparable physical processes rather than algorithmic constructs, the hybrid EDI neuro-patch highlights a new framework for self-organizing neural computation. This approach provides a pathway toward adaptive implants able of restoring and maintaining stable neural function through the intrinsic physics of matter with potential applications in epilepsy suppression, spinal injury bridging, and adaptive prosthetic control. Beyond its immediate clinical utility, a modular, scalable design could provide a platform for next-generation biohybrid or synthetic brain architectures, neuromorphic computation, and distributed neural networks with a clear pathway toward advanced brain–machine interfaces.