A Stable, Fast, and Fully Automatic Learning Algorithm for Predictive Coding Networks
This work addresses training challenges in neuroscience-inspired models, offering a more stable and efficient method for researchers in machine learning and computational neuroscience, though it is incremental as it modifies an existing approach.
The authors tackled the inefficiency and instability of training predictive coding networks by introducing a new temporal scheduling update rule, resulting in an algorithm (iPC) that consistently outperforms the original in test accuracy, efficiency, and convergence across image classification and language modeling benchmarks.
Predictive coding networks are neuroscience-inspired models with roots in both Bayesian statistics and neuroscience. Training such models, however, is quite inefficient and unstable. In this work, we show how by simply changing the temporal scheduling of the update rule for the synaptic weights leads to an algorithm that is much more efficient and stable than the original one, and has theoretical guarantees in terms of convergence. The proposed algorithm, that we call incremental predictive coding (iPC) is also more biologically plausible than the original one, as it it fully automatic. In an extensive set of experiments, we show that iPC constantly performs better than the original formulation on a large number of benchmarks for image classification, as well as for the training of both conditional and masked language models, in terms of test accuracy, efficiency, and convergence with respect to a large set of hyperparameters.