Speaker: Gabriel Stine (Jazayeri Lab) 

Title: Cerebellar adjustment of neocortical dynamics during non-motor learning

Abstract: The recurrent and interconnected architecture of the neocortex produces complex dynamics that give rise to cognitive behavior. However, this architecture poses a fundamental problem—the brain must control these complex dynamics to flexibly adapt cognitive behavior in ever-changing environments. In motor control, a similar problem is solved in part by the cerebellum, which builds internal models that predict the sensory consequences of actions, facilitating real-time corrections to movements and adaptability to environmental perturbations. My research asks whether the cerebellum, through its extensive closed-loop connections with neocortex, plays an analogous role in cognition by controlling neocortical dynamics that underlie non-motor behavior. In this talk, I’ll present our ongoing efforts to address this question using large-scale, simultaneous neuropixels recordings throughout the cerebello-thalamocortical pathway as monkeys perform a novel timing adaptation task. Preliminary findings suggest that the cerebellum conveys a control signal that adjusts, in real-time, the speed of ramping dynamics in neocortex, facilitating error correction of timing behavior. The results offer a first- glimpse into how the cerebellum uses internal models to control neocortical dynamics related to non-motor processes. 

Speaker: Iakovos Lazaridis (Graybiel Lab) 

Title: A dopamine-astrocyte pathway for behavioral flexibility

Abstract: Behavioral flexibility depends on adapting actions as internal state and environmental demands change, a process strongly shaped by dopamine. How dopaminergic signals are translated into shifts in behavioral state in the dorsal striatum remains unclear. Here we show that astrocytes act as a state-sensitive interface linking dopamine signaling to motivational state. Using dual-color photometry during reinforcement-guided behavior, we find that astrocytic Ca²⁺ activity does not encode specific choices or outcomes, but instead peaks around transitions from disengaged to engaged states. Dopamine fluctuations typically precede astrocyte events, consistent with dopamine recruiting astrocytes during state change. Astrocyte signals preferentially decode behavioral state over trial outcome. Dopamine-evoked astrocyte Ca²⁺ responses require astrocytic Drd2, and astrocyte activation suppresses local dopamine and drives astrocyte-dependent adenosine release. Together, these findings identify astrocytes as state-sensitive integrators of dopaminergic input that regulate dopamine availability and support behavioral flexibility.