AI & ML Efficiency Breakthrough

Introduces a stable backpropagation-free training framework for physical and photonic neural networks.

March 27, 2026

Original Paper

Local learning for stable backpropagation-free neural network training towards physical learning

Yaqi Guo, Fabian Braun, Bastiaan Ketelaar, Stephanie Tan, Richard Norte, Siddhant Kumar

arXiv · 2603.24790

The Takeaway

By using local learning and forward-only evaluations, this method removes the need for digital automatic differentiation, enabling energy-efficient, in-situ training on physical hardware that cannot easily compute gradients.

From the abstract

While backpropagation and automatic differentiation have driven deep learning's success, the physical limits of chip manufacturing and rising environmental costs of deep learning motivate alternative learning paradigms such as physical neural networks. However, most existing physical neural networks still rely on digital computing for training, largely because backpropagation and automatic differentiation are difficult to realize in physical systems. We introduce FFzero, a forward-only learning