An evolutionary computational theory of prefrontal executive function in decision-making

Area of Science:

• Neuroscience
• Cognitive Science
• Evolutionary Biology

Background:

• The prefrontal cortex is crucial for executive control and decision-making.
• Adaptive behavior relies on balancing exploration of new strategies and exploitation of known ones.

Purpose of the Study:

• To present a computational theory on the evolutionary trajectory of the prefrontal cortex.
• To explain the gradual addition of Bayesian inferential capabilities for optimizing decision-making.

Main Methods:

• Computational modeling of prefrontal cortex evolution from rodents to humans.
• Analysis of inferential steps in decision arbitration: factual reactive, factual proactive, and counterfactual inferences.
• Integration of model-free and model-based reinforcement learning (RL).

Main Results:

• Identified three stages of inference evolution: rodents (factual reactive), primates (factual proactive), and humans (counterfactual).
• Clarified the integration of RL models via 'strategy creation'.
• Demonstrated that human counterfactual inferences equate to hypothesis testing for adaptive processes.

Conclusions:

• The evolution of the prefrontal cortex involves progressive enhancement of Bayesian inferential abilities.
• Counterfactual reasoning and hypothesis testing provide humans with a significant evolutionary advantage in adaptive behavior.
• The theory offers a unified account of decision-making, strategy learning, and RL integration.