Weight Feedback Computes the Jacobian Transpose Locally in Modern Deep Networks

Jul 16, 2026

A new preprint demonstrates that predictive coding (PC) can avoid the non-local Jacobian-transpose operation by factoring it into three locally available terms for layers with frozen normalization. The resulting method, WF-Act-PC, removes the need for autograd backward passes in error transport and, on benchmarks like CIFAR-10 and Tiny-ImageNet, matches or exceeds backpropagation performance on deeper architectures, outperforming previous PC methods.

Why it matters: This work addresses a longstanding obstacle to biologically plausible learning by eliminating a key non-local operation in predictive coding, narrowing the performance gap with backpropagation in deep networks.

Full story at: arXiv Machine Learning ↗