Latching chains in K-nearest-neighbor and modular small-world networks.

Area of Science:

• Computational Neuroscience
• Network Science

Background:

• Neural adaptation and pattern correlation enable latching dynamics for retrieving sequential information.
• A modular latching chain model was previously proposed to explain structured brain transitions.
• Cortical areas exhibit diverse network structures influencing neural processing.

Purpose of the Study:

• To investigate the impact of structural parameters (rewiring probability, threshold, noise, feedback) on latching dynamics.
• To compare latching dynamics in K-nearest-neighbor and modular networks with modular structures.
• To analyze how network architecture affects sequential pattern retrieval in neural systems.

Main Methods:

• Simulated two network connection schemes: K-nearest-neighbor and modular networks.
• Varied structural parameters including rewiring probability, threshold, noise, and feedback connections.
• Measured latching chains using intra-modular chain length and inter-modular association transitions.

Main Results:

• Both network types showed similar phase transitions with decreasing threshold and rewiring probability.
• Modular networks demonstrated enhanced latching within the small-world connectivity range.
• K-nearest-neighbor networks were more robust to rewiring changes, while modular networks were more robust to noise and feedback.

Conclusions:

• Network structure significantly modulates latching dynamics and sequential information processing.
• Modular networks offer advantages in specific connectivity regimes for pattern retrieval.
• Findings provide insights into the relationship between neural network architecture and cognitive functions.