Sakana AI proposes backpropagation alternative for very deep nets

Sakana AI researchers introduced PC-ALM (Augmented Lagrangian Predictive Coding), a layer-local training method that recovers backpropagation-level gradients without a global backward pass. It matches standard backprop across network widths and depths from 8 to 128 layers, and trains 1,000-layer residual networks within about 2 percentage points of backprop accuracy on MNIST. The team released MIT-licensed JAX code.

Why it matters: Backpropagation's need for global, sequential gradient computation is a real bottleneck for scaling and parallelizing training; a workable layer-local alternative could eventually enable more efficient or hardware-friendly training pipelines. Sakana has been prolific this year, including its recent Fugu multi-agent models, positioning itself as a research-heavy lab distinct from the big labs' product race.