A refreshable memristor leveraging dynamic ferro-ionic phase allocation enables neural reuse, improving classification accuracy, reducing consumption, and accelerating training speed compared to randomly initialized models, enhancing efficiency and adaptability in neuromorphic hardware.
- Jiangang Chen
- Zhixing Wen
- Fucai Liu
