A neural compass that combines flexibility and stability

Theoretical studies have long suggested that internal representations most prominently associated with mammalian navigation -such as head direction cells, grid cells, and place cells- are maintained by attractor dynamics. The size and complexity of the mammalian brain, however, has made it difficult to test models that seek to explain how such network dynamics are generated and used for navigational behavior. We study attractor dynamics in the fly, Drosophila melanogaster, which displays many navigational behaviors, including path integration and place learning. We are combining physiology, optogenetics, and behavior with electron microscopic circuit reconstruction and theoretical modeling to understand the cellular and circuit basis of such recurrent circuit dynamics in the context of adaptive navigation. In this talk, I will focus on how inhibitory Hebbian plasticity enables the fly's internal compass to function in diverse settings by flexibly transforming visual cues into a stable internal representation of heading.