Abstract:Brain's connectivity reveals modular and hierarchical structures at various scales. This organization is typically believed to support the coexistence of segregation (specialization) and integration (binding) of information. Motivated by developmental processes, some authors have studied the self-organization of neural networks into modular hierarchies mediated by adaptive mechanism under spontaneous neural activity. Following evidence that the sensory cortices organize into assemblies under selective stimuli, other authors have shown that stable neural assemblies can emerge in random neural networks due to targeted stimulation, embedding various forms of synaptic plasticity in presence of homeostatic and/or control mechanisms. Here, we show that simple spike-timing-dependent plasticity (STDP) rules, based only on pre- and post-synaptic spike times, can also lead to the stable encoding of memories in the absence of any control mechanism. We develop a model of spiking neurons, trained to stimuli targeting different sub-populations. The model is intended to satisfy biologically plausible features: (i) it contains excitatory and inhibitory neurons with Hebbian and anti-Hebbian STDP; (ii) neither the neuronal activity nor the synaptic weights are frozen after the learning phase. Instead, the neurons are allowed to fire spontaneously while synaptic plasticity remains active. We find that only the combination of two inhibitory STDP sub-populations allows for the formation of stable modular organization in the network, with each sub-population playing a distinctive role. The Hebbian sub-population controls for the firing activity, while the anti-Hebbian one promotes pattern selectivity. After the learning phase, the network settles into an asynchronous irregular resting-state. This post-learning activity is associated with spontaneous memory recalls, which turn fundamental for the long-term consolidation of the learned memories. Due to its simplicity, the model, here introduced, can represent a test-bed for further investigations on the role played by STDP on memory storing and maintenance.

Emergence and reconfiguration of modular structure for synaptic neural networks during continual familiarity detection

Emergence and reconfiguration of modular structure for artificial neural networks during continual familiarity detection

Emergence and long-term maintenance of modularity in plastic networks of spiking neurons

Emergence and maintenance of modularity in neural networks with Hebbian and anti-Hebbian inhibitory STDP

Modularity Facilitates Classification Performance of Spiking Neural Networks for Decoding Cortical Spike Trains

Meta-learning synaptic plasticity and memory addressing for continual familiarity detection

Beyond modularity: Fine-scale mechanisms and rules for brain network reconfiguration

Modular Growth of Hierarchical Networks: Efficient, General, and Robust Curriculum Learning

Modular representations emerge in neural networks trained to perform context-dependent tasks

Synergistic pathways of modulation enable robust task packing within neural dynamics

A Study of Biologically Plausible Neural Network: The Role and Interactions of Brain-Inspired Mechanisms in Continual Learning

Modular Continual Learning in a Unified Visual Environment

Optimal modularity and memory capacity of neural reservoirs

Incremental Self-Organization of Spatio-Temporal Spike Pattern Detection

Don't Cut Corners: Exact Conditions for Modularity in Biologically Inspired Representations

Concurrence of form and function in developing networks and its role in synaptic pruning

Flexible multitask computation in recurrent networks utilizes shared dynamical motifs

Co-evolution of Functional Brain Network at Multiple Scales during Early Infancy

Growing Brains: Co-emergence of Anatomical and Functional Modularity in Recurrent Neural Networks

Emergence of cortical network motifs for short-term memory during learning

Incorporating neuro-inspired adaptability for continual learning in artificial intelligence