A Programmer-Interpreter Neural Network Architecture for Prefrontal Cognitive Control.

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Neuroscience

Background:

• The prefrontal cortex (PFC) is known to exert cognitive control by biasing information processing and modulating response selection.
• This control is often conceptualized as top-down influence within a cortical hierarchy involving prefrontal-basal ganglia loops.

Purpose of the Study:

• To propose a new computational framework for understanding how the PFC flexibly biases sensorimotor patterns based on goals and context.
• To introduce an interpreter-programmer scheme where the PFC acts as a programmer modulating a reusable neural network (interpreter).

Main Methods:

• Developed a computational model based on the programmer-interpreter scheme.
• Validated the model using the 1-2-AX task, a standard test for executive function.

Main Results:

• The proposed computational framework demonstrates robust, scalable, and flexible goal-directed behavior.
• The PFC, as a programmer, effectively biases sensorimotor selection by providing input parameters without altering network connectivity.

Conclusions:

• The programmer-interpreter scheme offers a plausible mechanism for prefrontal-(pre)motor cortical hierarchies.
• This framework supports the flexible realization of multiple goals through iterative application across hierarchical layers.