We present a neuronal network model of Basal Ganglia that departs from the classical Go/NoGo picture of the function of its key pathways - the direct and indirect pathways (DP & IP). The model is instantiated in a simple action selection task. Striatal dopamine is assumed to switch between DP and IP activation. Simulations reveal that between the Go and NoGo regimes, exhibited at extreme values of dopamine, the system displays, at intermediate values of dopamine, a new Explore regime, which enables it to explore the space of action alternatives. The exploratory dynamics originates from the chaotic dynamics of pallido-subthalamic loop. Following the tradition of applying reinforcement learning (RL) concepts to BG function, we associate this new regime with the Explorer, a key component in Actor-Critic framework. © 2011 IEEE.
View more info for "Expanding the Go/NoGo depiction of the action of Basal Ganglia pathways"
|Proceedings of the International Joint Conference on Neural Networks