TIDE: Asymmetric Neural Circuits for Stabilized Temporal Inhibitory-Excitatory Dynamics

2026-05-19

AI Summary

This work addresses the lack of neural dynamical stability in existing Continuous-Time Models (CTMs), which struggle to balance biological plausibility with learning efficiency. The authors propose TIDE, a novel architecture that, for the first time, integrates network stability theory, Dale’s law, and the biologically observed 80:20 excitatory-inhibitory (E-I) ratio constraint into an end-to-end trainable model. TIDE achieves stable neural dynamics through asymmetric E-I circuits, Wilson–Cowan dynamics, and lateral inhibition, while incorporating a game-theoretic loss to optimize its energy landscape. Empirically, TIDE improves average top-1 accuracy by 1.65% on ImageNet while requiring less than half the training time of conventional CTMs. The framework further provides rigorous theoretical guarantees regarding convergence, dynamical stability, and computational complexity.

Abstract

Recent Continuous Thought Machine architecture decouples internal computation from external inputs via neural dynamics, but relies on multi-layer perceptrons without stability guarantees. We propose to model neural dynamics using asymmetric Excitatory-Inhibitory (E-I) networks, which can be stabilized via principles from network theory and can be expressed as energy-based systems optimized through a game-theoretic loss. Building on this perspective, we introduce Temporal Inhibitory-Excitatory Dynamic Engine (TIDE), a neuro-inspired architecture that computes internal representations through neural dynamics stabilized by incorporating the Wilson-Cowan dynamics and lateral inhibition. TIDE balances biological realism by, for instance, using Hierarchical Receptive Fields and enforcing Dale's principle to ensure a realistic $80:20$ E-I balance ratio with an end-to-end trainable architecture. The aim of this paper is to introduce a new architecture that brings neuro-inspired learning to the forefront. We present proofs of convergence, stability, and complexity bounds, along with empirical ablation studies. Overall, TIDE surpasses CTM with under $50\%$ of the training time and improves $\texttt{top-1}$ accuracy by an average of $+1.65\%$ on ImageNet under various perturbations.

Problem

• neural dynamics
• stability
• excitatory-inhibitory balance
• biological plausibility
• continuous thought machine

Innovation

• asymmetric E-I networks
• neural dynamics stabilization
• energy-based models
• Wilson-Cowan dynamics
• Dale's principle