Choice selective inhibition drives stability and competition in decision circuits

Area of Science:

• Neuroscience
• Computational Neuroscience
• Decision Neuroscience

Background:

• Cortical neuron firing rates encode choices during perceptual decision-making.
• Both excitatory and inhibitory neurons exhibit choice selectivity.
• The functional role of inhibitory choice selectivity in decision circuits remains unclear.

Purpose of the Study:

• To investigate the functional consequences of inhibitory choice selectivity in decision-making circuits.
• To model how inhibitory neuron specificity impacts circuit dynamics and decision outcomes.
• To explore the interplay between inhibitory specificity and recurrent excitation.

Main Methods:

• Development of a circuit model of decision-making incorporating specific inhibitory inputs and outputs.
• Analysis of model dynamics under different connectivity motifs (competitive, feedback).
• Comparison with recurrent neural networks trained for decision tasks.

Main Results:

• Selective inhibition expands the range of circuit parameters supporting decision-making.
• Inhibitory specificity modulates recurrent excitation strength and circuit stability.
• Decision speed-accuracy trade-offs are determined by inhibitory output specificity and circuit dynamics (saddle-point stabilization/destabilization).
• Recurrent neural networks show similar dependencies on inhibitory specificity and excitation strength.

Conclusions:

• Selective inhibition plays dual roles: stabilizing excitatory populations and enhancing competition.
• Inhibitory neuron specificity is critical for tuning decision-making circuit function.
• The findings provide insights into neural mechanisms underlying perceptual decisions and their speed-accuracy dynamics.