Heterogeneous Time Constants Improve Stability in Equilibrium Propagation

2026-03-03

AI Summary

This work addresses a key limitation in existing Equilibrium Propagation (EP) methods, which employ uniform time steps and thereby neglect the heterogeneity of neuronal membrane time constants observed in biological systems, leading to constrained training stability. To bridge this gap, the study introduces biologically grounded heterogeneous time constants into the EP framework for the first time. Specifically, each neuron is assigned a unique time constant sampled from a neuroscience-inspired distribution, yielding a heterogeneous time-stepping mechanism. This approach not only preserves task performance on par with current methods but also substantially enhances training stability, effectively reconciling biological plausibility with algorithmic robustness.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Bio-inspired Learning
• Search and Optimization: Learning to Search

Application Category

• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Equilibrium propagation (EP) is a biologically plausible alternative to backpropagation for training neural networks. However, existing EP models use a uniform scalar time step dt, which corresponds biologically to a membrane time constant that is heterogeneous across neurons. Here, we introduce heterogeneous time steps (HTS) for EP by assigning neuron-specific time constants drawn from biologically motivated distributions. We show that HTS improves training stability while maintaining competitive task performance. These results suggest that incorporating heterogeneous temporal dynamics enhances both the biological realism and robustness of equilibrium propagation.

Problem

• Equilibrium Propagation
• heterogeneous time constants
• training stability
• biological plausibility
• neural networks

Innovation

• Heterogeneous Time Constants
• Equilibrium Propagation
• Biological Plausibility
• Training Stability
• Neuron-specific Dynamics