Sparse Coding and Temporal Pattern Learning Co-Mediated by Dual Spike-Timing-Dependent Plasticity in a Multilayer...

Area of Science:

• Computational neuroscience
• Neural network modeling
• Synaptic plasticity

Background:

• Excitatory-inhibitory (E-I) circuits are crucial for brain function, but their multilayer learning dynamics are not fully understood.
• Spiking neural networks (SNNs) offer a biologically plausible framework for studying neural computation.

Purpose of the Study:

• To investigate the cooperative learning dynamics of E-I circuits in a multilayer feedforward SNN.
• To examine the roles of excitatory and inhibitory spike-timing-dependent plasticity (eSTDP and iSTDP) in shaping network function.

Main Methods:

• Constructed a multilayer feedforward SNN using Izhikevich neurons (regular spiking and fast spiking cells).
• Implemented intra-layer E-I connectivity and simulated eSTDP and iSTDP.
• Performed parameter grid scans and analyzed network outputs for sparse coding and temporal pattern learning.

Main Results:

• Fast spiking (FS) cell-mediated inhibition counteracts adaptation biases, promoting weight differentiation and stable learning.
• iSTDP expands the parameter space for stable E-I cooperative learning.
• The network demonstrated enhanced sparse coding and improved temporal pattern recognition (d' increased ~1.90x without FS inhibition).
• FS circuits maintained higher stimulus-related information in deeper layers.

Conclusions:

• Cooperative plasticity between excitatory and inhibitory neurons is essential for robust neural coding and learning in multilayer SNNs.
• Adaptive inhibitory regulation by FS cells enhances network stability, information processing, and temporal pattern learning.
• Findings inform the design of neuromorphic systems with adaptive inhibitory mechanisms.