The alteration in microstructure, induced by ion migration due to applied voltage, constitutes a pivotal factor influencing the performance of the memristor. This phenomenon adversely impacts the stability of the memristor, posing challenges for its practical applications. Notably, the defects present in oxide films, serving as the functional layer in the memristor, assume a crucial role in determining the stability of the artificial synapse—a fundamental component of neuromorphic computing. The precise regulation of defect distribution and density at the nanoscale by growing films directly poses a formidable challenge. In this investigation, a memristor composed of strontium titanate (SrTiO3) was fabricated, exhibiting improved stability in resistive switching during I–V cycles and enhanced multilevel storage performance through the implementation of Au ions implantation. Furthermore, these devices were simulated as neural synapses and integrated into artificial neural networks. A comprehensive array of characterizations was executed to scrutinize the microscopic effects of ion implantation. This involved analyzing changes in elemental composition, structural damage, and spectral characteristics of the films. These findings offer a viable strategy for enhancing the resistive switching performance of oxide thin film devices through the judicious application of ion implantation.