Benchmarking Predictive Coding Networks – Made Simple

Abstract

In this work, we tackle the problems of efficiency and scalability for predictive coding networks in machine learning. To do so, we first propose a library called PCX, whose focus lies on performance and simplicity, and provides a user-friendly, deep-learning oriented interface. Second, we use PCX to implement a large set of benchmarks for the community to use for their experiments. As most works propose their own tasks and architectures, do not compare one against each other, and focus on small-scale tasks, a simple and fast open-source library adopted by the whole community would address all of these concerns. Third, we perform extensive benchmarks using multiple algorithms, setting new state-of-the-art results in multiple tasks and datasets, as well as highlighting limitations inherent to PC that should be addressed. Thanks to the efficiency of PCX, we are able to analyze larger architectures than commonly used, providing baselines to galvanize community efforts towards one of the main open problems in the field: scalability.

Download paper here

BibTex

@misc{pinchetti2024benchmarkingpredictivecodingnetworks, title={Benchmarking Predictive Coding Networks – Made Simple}, author={Luca Pinchetti and Chang Qi and Oleh Lokshyn and Gaspard Olivers and Cornelius Emde and Mufeng Tang and Amine M’Charrak and Simon Frieder and Bayar Menzat and Rafal Bogacz and Thomas Lukasiewicz and Tommaso Salvatori}, year={2024}, eprint={2407.01163}, archivePrefix={arXiv}, primaryClass={cs.LG}, url={https://arxiv.org/abs/2407.01163}, }