Layer Collaboration in the Forward-Forward Algorithm

TL;DR

This paper investigates the limitations of the forward-forward algorithm's layer independence, proposing an improved collaborative version that enhances information flow and learning capacity without extra computational costs.

Contribution

It introduces a novel layer collaboration mechanism for the forward-forward algorithm, improving its optimization process and theoretical understanding.

Findings

Enhanced information flow between layers.

Improved objective metrics in experiments.

Theoretically motivated by functional entropy theory.

Abstract

Backpropagation, which uses the chain rule, is the de-facto standard algorithm for optimizing neural networks nowadays. Recently, Hinton (2022) proposed the forward-forward algorithm, a promising alternative that optimizes neural nets layer-by-layer, without propagating gradients throughout the network. Although such an approach has several advantages over back-propagation and shows promising results, the fact that each layer is being trained independently limits the optimization process. Specifically, it prevents the network's layers from collaborating to learn complex and rich features. In this work, we study layer collaboration in the forward-forward algorithm. We show that the current version of the forward-forward algorithm is suboptimal when considering information flow in the network, resulting in a lack of collaboration between layers of the network. We propose an improved version that supports layer collaboration to better utilize the network structure, while not requiring any additional assumptions or computations. We empirically demonstrate the efficacy of the proposed version when considering both information flow and objective metrics. Additionally, we provide a theoretical motivation for the proposed method, inspired by functional entropy theory.