Integrated Ising model with global inhibition for decision-making.

Area of Science:

• Computational Neuroscience
• Decision-Making Models
• Animal Behavior

Background:

• The drift-diffusion model is a primary framework for decision-making, but recent findings challenge its explanatory power.
• Increased inhibitory tone during difficult discrimination tasks is observed, yet its neural origin remains unclear.
• Existing models do not fully capture the complexities of real-world decision-making, especially concerning inhibition.

Purpose of the Study:

• To extend existing directional decision-making models by incorporating global inhibition.
• To develop an integrated Ising-type model for two-choice decision-making.
• To explain how neural inhibition can improve decision-making accuracy and explore the brain's operational regime.

Main Methods:

• Developed an integrated Ising-type model incorporating global inhibition for two-choice decision tasks.
• Extended a recently developed model for directional decision-making in animals moving in real space.
• Compared model predictions with experimental results to validate the proposed mechanism.

Main Results:

• The proposed Ising-type model successfully explains how global inhibition can enhance decision-making accuracy.
• Model simulations suggest the brain's decision-making activity operates near a critical transition between ordered and disordered phases.
• This proximity to a critical region exhibits unique dynamics beneficial for decision processes.

Conclusions:

• The integrated Ising-type model provides a novel framework for understanding the role of inhibition in decision-making.
• The brain's operation near a critical transition point offers computational advantages for maximizing reward and minimizing cost.
• This research sheds light on the neural mechanisms underlying efficient decision-making in complex environments.