Self-tuned critical anti-Hebbian networks

Area of Science:

• Neuroscience
• Dynamical Systems Theory
• Computational Biology

Background:

• Maintaining a balance between neural excitation and inhibition is crucial for nervous system function.
• Global strategies for achieving this balance often result in a mix of stable and near-unstable modes.

Purpose of the Study:

• To present a simple abstract model demonstrating local strategies for achieving excitation-inhibition balance.
• To investigate the dynamical properties and emergent behaviors of such a system.

Main Methods:

• Development of an abstract model employing local "anti-Hebbian" evolution rules.
• Analysis of the system's dynamics at long time scales.

Main Results:

• All degrees of freedom achieve a critical state, balancing excitation and inhibition.
• Complex "breakout" dynamics emerge, with modes oscillating between prominence and extinction.
• The model exhibits long-tailed statistical behaviors, suggesting self-organized criticality.

Conclusions:

• Local, anti-Hebbian-based strategies can lead to critical dynamics in neural systems.
• This emergent criticality may underlie complex behaviors and long-tailed statistics observed in biological systems.
• The model provides a framework for understanding self-organized critical states in neural networks.