Toward Lifelong Learning in Equilibrium Propagation: Sleep-like and Awake Rehearsal for Enhanced Stability

2025-08-12

AI Summary

To address catastrophic forgetting in Equilibrium Propagation (EP)-trained recurrent neural networks (RNNs) during continual learning, this paper proposes Sleep-like Replay Consolidation (SRC), a biologically inspired algorithm modeled on human sleep-dependent memory consolidation. SRC introduces, for the first time, a biologically plausible sleep-like replay mechanism into the EP-RNN framework, synergistically combined with wake-phase replay to enable gradient-free, biologically interpretable continual learning—without backpropagation through time (BPTT). By periodically reactivating latent representations of previously learned tasks in offline replay phases, SRC stabilizes synaptic weights and mitigates representational interference. Experiments on MNIST and Fashion-MNIST demonstrate that EP-RNN+SRC achieves long-term task retention comparable to or exceeding BPTT-based baselines, without gradient clipping or external memory buffers, thereby effectively balancing plasticity and stability in continual learning.

Technology Category

• Machine Learning: Life-Long and Continual Learning
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Social networks and social learning
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Recurrent neural networks (RNNs) trained using Equilibrium Propagation (EP), a biologically plausible training algorithm, have demonstrated strong performance in various tasks such as image classification and reinforcement learning. However, these networks face a critical challenge in continuous learning: catastrophic forgetting, where previously acquired knowledge is overwritten when new tasks are learned. This limitation contrasts with the human brain's ability to retain and integrate both old and new knowledge, aided by processes like memory consolidation during sleep through the replay of learned information. To address this challenge in RNNs, here we propose a sleep-like replay consolidation (SRC) algorithm for EP-trained RNNs. We found that SRC significantly improves RNN's resilience to catastrophic forgetting in continuous learning scenarios. In class-incremental learning with SRC implemented after each new task training, the EP-trained multilayer RNN model (MRNN-EP) performed significantly better compared to feedforward networks incorporating several well-established regularization techniques. The MRNN-EP performed on par with MRNN trained using Backpropagation Through Time (BPTT) when both were equipped with SRC on MNIST data and surpassed BPTT-based models on the Fashion MNIST, Kuzushiji-MNIST, CIFAR10, and ImageNet datasets. Combining SRC with rehearsal, also known as "awake replay", further boosted the network's ability to retain long-term knowledge while continuing to learn new tasks. Our study reveals the applicability of sleep-like replay techniques to RNNs and highlights the potential for integrating human-like learning behaviors into artificial neural networks (ANNs).

Problem

• Addresses catastrophic forgetting in recurrent neural networks during continuous learning
• Proposes sleep-like replay consolidation to enhance knowledge retention in Equilibrium Propagation
• Combines sleep and awake rehearsal for improved lifelong learning stability

Innovation

• Sleep-like replay consolidation algorithm for RNNs
• Combining sleep-like and awake rehearsal techniques
• Equilibrium Propagation with enhanced memory retention