AI Insight
This study presents a new learning algorithm for spiking neural networks that solves the temporal credit assignment problem using local spike timing information, rather than relying on backpropagation approximations. The gradient tunneling (GT) algorithm reformulates credit assignment as a state separation problem, extracting task-relevant information from current neural states based on historical activity patterns. The approach demonstrates strong performance on long-timescale tasks and memory retention while using fewer parameters than existing methods, and bridges biological plausibility with computational efficiency.
Why it matters
This work addresses a fundamental challenge in understanding how the brain learns from temporally delayed feedback, potentially offering insights into biological learning mechanisms. The algorithm could enable more efficient neuromorphic computing systems that better mimic brain-like computation while maintaining compatibility with existing artificial neural network architectures.
Understand the Science
⚠️ Preprint – Noch nicht peer-reviewed
Dieser Artikel wurde noch nicht von unabhängigen Experten begutachtet. Die Ergebnisse sind vorläufig und sollten mit Vorsicht interpretiert werden.
-cross
Abstract: The brain uses discrete spikes for dynamic computation, yet, how neural microcircuits (NMCs) solve temporal credit assignment using local spike timing remains a fundamental open question. Dominant spiking neural network (SNN) approaches circumvent this by approximating backpropagation through surrogate gradients, decoupling learning from biological spike timing. Here, we reformulate temporal credit assignment as a state separation problem: extracting task-required components induced by historical perturbations directly from the current neural state. This enables an online feedback learning framework for NMCs through a gradient tunneling (GT) algorithm and the lead-lag expansion technique that derives credit assignment from local synaptic spike timing, while remaining compatible with ANN-SNN hybrid architectures. Experimentally, GT-trained NMCs excel at long-timescale evidence integration and noise-robust memory retention, and perform comparably to leading SNN online learning methods on real-world benchmarks with far fewer parameters. The proposed framework addresses the two-decade-old NMC feedback learning problem and suggests a computationally plausible explanation for the brain’s learning mechanisms.