The basal ganglia: Summary and future modeling resxsearch

Abstract

The pivotal idea of the Go-Explore-NoGo (GEN) approach to BG modeling, expounded in this book, is the hypothesis that the complex dynamics of the STN–GPe loop can introduce certain randomness in the action selection mechanisms that occur downstream in other BG areas as well as the cortex, thalamus, and other subcortical structures. The indirect pathway, that consists of the STN–GPe loop, in addition to its classical role as a source of movement inhibition, also serves as a source of randomness, and therefore, in the jargon of reinforcement learning, can control the levels of exploration in action selection. In this chapter, we summarize how the Go-Explore-NoGo approach can account for several functions of the basal ganglia. We also provide a few ideas on how to use the Go-Explore-NoGo approach to simulate behavioral performance in other basal ganglia-related functions (e.g., attention, working memory, episodic memory) and disorders (e.g., schizophrenia).