Deviance detection via competitive inhibition between local neocortical ensembles

Area of Science:

• Neuroscience
• Computational Neuroscience
• Systems Neuroscience

Background:

• Deviance detection (DD) traditionally attributed to specialized neurons or brain regions.
• Neocortical circuits process unexpected stimulus changes.
• Predictive sensory processing in the Neocortex remains incompletely understood.

Purpose of the Study:

• Propose an intrinsic, local network-level theory for deviance detection (DD).
• Investigate how local interactions generate deviance-driven neural dynamics.
• Explore a novel inhibitory connectivity motif for DD.

Main Methods:

• Developed two computational models with varying biophysical abstraction.
• Implemented a competitive inhibition motif with selective (dis)inhibition.
• Simulated neural tuning shifts observed during experimental DD.

Main Results:

• Demonstrated that local network interactions can generate DD phenomena.
• Showed ensemble priming via competitive inhibition encodes short-term memory.
• Identified a local mechanism for deviance-driven stimulus representation shifts.

Conclusions:

• Deviance detection emerges intrinsically from local neocortical network dynamics.
• Competitive inhibition and selective (dis)inhibition provide a local mechanism for DD.
• This theory resolves confounding aspects of predictive sensory processing and offers testable predictions.