Lock-in EP: An In-Situ Training Algorithm for Oscillatory Hardware

Excerpt

arXiv:2610.07283v1 Announce Type: new Abstract: Analog hardware platforms offer the potential to reduce energy consumption over digital architectures, but in order to succeed, large-scale analog systems must also be able to operate with or recover from the variability of their components. Towards this goal, we derive and demonstrate the lock-in equilibrium propagation (LIEP) training method. LIEP provides local gradient information for each component in an oscillatory network without separate fo