HCL-FF: Hierarchical and Contrastive Learning for Forward-Forward Algorithm

2026-05-23

AI Summary

This work addresses the limitations of the Forward-Forward (FF) algorithm—namely, its lack of inter-layer coordination and semantic ambiguity—which hinder both performance and interpretability. The authors propose a bottom-up hierarchical forward-forward training mechanism combined with a decoupled supervised contrastive alignment objective, enabling, for the first time within the FF framework, semantically explicit and cross-layer collaborative local learning. By progressively refining representations from coarse to fine and employing localized “goodness” metrics, the method substantially enhances model expressiveness. Empirical results demonstrate significant accuracy improvements of 5.46%, 17.00%, and 12.51% on CIFAR-10, CIFAR-100, and Tiny-ImageNet, respectively, establishing a new state-of-the-art among FF-based approaches.

Abstract

Deep neural networks trained with backpropagation have achieved outstanding performance in vision tasks but remain biologically implausible, computationally demanding, and difficult to interpret. The Forward-Forward (FF) algorithm offers a promising alternative by training each layer independently through local goodness objectives. However, its purely local optimization lacks hierarchical coordination across layers, and the decoupling of goodness from features leaves the representations unconstrained and semantically ambiguous. We propose a Hierarchical and Contrastive Learning FF framework (HCL-FF) to address these limitations. HCL-FF introduces (1) a coarse-to-fine hierarchical learning strategy that guides representations from low-level cues to high-level semantics, and (2) a supervised contrastive objective that enforces class-discriminative alignment after goodness decoupling. Experiments on CIFAR-10, CIFAR-100, and Tiny-ImageNet demonstrate that HCL-FF achieves new state-of-the-art performance among FF-based methods, with notable accuracy gains of +5.46%, +17.00%, and +12.51%, respectively.

Problem

• Forward-Forward algorithm
• hierarchical coordination
• representation ambiguity
• local optimization
• semantic clarity

Innovation

• Forward-Forward algorithm
• hierarchical learning
• contrastive learning
• biologically plausible learning
• local optimization