Mixed-signal implementation of feedback-control optimizer for single-layer Spiking Neural Networks
This work addresses the problem of enabling expressive, on-chip training for scalable and adaptive neuromorphic computing, representing an incremental co-design approach.
The authors tackled the challenge of on-chip learning for neuromorphic systems by implementing a feedback-control optimizer on a mixed-signal neuromorphic processor, demonstrating that it matches the performance of numerical simulations and gradient-based baselines in tasks like binary classification and the Yin-Yang problem.
On-chip learning is key to scalable and adaptive neuromorphic systems, yet existing training methods are either difficult to implement in hardware or overly restrictive. However, recent studies show that feedback-control optimizers can enable expressive, on-chip training of neuromorphic devices. In this work, we present a proof-of-concept implementation of such feedback-control optimizers on a mixed-signal neuromorphic processor. We assess the proposed approach in an In-The-Loop(ITL) training setup on both a binary classification task and the nonlinear Yin-Yang problem, demonstrating on-chip training that matches the performance of numerical simulations and gradient-based baselines. Our results highlight the feasibility of feedback-driven, online learning under realistic mixed-signal constraints, and represent a co-design approach toward embedding such rules directly in silicon for autonomous and adaptive neuromorphic computing.