Evolutionary game theory and the evolution of neuron populations, firing rates, and decision making

Recently, there have been calls to apply evolutionary game theory to neuroscience. Ours, to the best of our knowledge, is the first application of dynamical evolutionary games to decision making in neuroscience. Firing neurons are the players in the game. The strategy of the players is their firing rate. Neurons with equal firing rates define a population. The neurons (players) do not know the rules of the game, they do not know what the reward is, they are not required to be rational and they do not even know they are playing the game. The interactions among the neurons is inhibitory. The theory confirms experimental data about decision making in vision (to move the eyes or not to a known target) in that: (i) A parameter of the game model (variance in the size of the neuron populations that participate in the game) determines how many distinct populations of firing neurons participate in the decision to move the eye to a target; (ii) if the populations of firing neurons are localized, then the solution of the game dictates how many loci in the brain participate in the decision to move the eye; (iii) the theory clarifies the difference between ultimate (evolutionary) factors and their effect on proximate (immediate) factors that influence decision making in the brain; (iv) as expected, the theory predicts that quick decisions are associated with more errors and slow decision are associated with fewer errors—however, the theory ties these factors to the evolution of brain decision making processes.