Search for: All records

Abstract The cerebellum regulates nonmotor behavior, but the routes of influence are not well characterized. Here we report a necessary role for the posterior cerebellum in guiding a reversal learning task through a network of diencephalic and neocortical structures, and in flexibility of free behavior. After chemogenetic inhibition of lobule VI vermis or hemispheric crus I Purkinje cells, mice could learn a water Y-maze but were impaired in ability to reverse their initial choice. To map targets of perturbation, we imaged c-Fos activation in cleared whole brains using light-sheet microscopy. Reversal learning activated diencephalic and associative neocortical regions. Distinctive subsets of structures were altered by perturbation of lobule VI (including thalamus and habenula) and crus I (including hypothalamus and prelimbic/orbital cortex), and both perturbations influenced anterior cingulate and infralimbic cortex. To identify functional networks, we used correlated variation in c-Fos activation within each group. Lobule VI inactivation weakened within-thalamus correlations, while crus I inactivation divided neocortical activity into sensorimotor and associative subnetworks. In both groups, high-throughput automated analysis of whole-body movement revealed deficiencies in across-day behavioral habituation to an open-field environment. Taken together, these experiments reveal brainwide systems for cerebellar influence that affect multiple flexible responses. 

Abstract—We present the Seldonian Toolkit, which enables software engineers to integrate provably safe and fair machine learning algorithms into their systems. Software systems that use data and machine learning are routinely deployed in a wide range of settings from medical applications, autonomous vehicles, the criminal justice system, and hiring processes. These systems, however, can produce unsafe and unfair behavior, such as suggesting potentially fatal medical treatments, making racist or sexist predictions, or facilitating radicalization and polarization. To reduce these undesirable behaviors, software engineers need the ability to easily integrate their machine- learning-based systems with domain-specific safety and fairness requirements defined by domain experts, such as doctors and hiring managers. The Seldonian Toolkit provides special machine learning algorithms that enable software engineers to incorporate such expert-defined requirements of safety and fairness into their systems, while provably guaranteeing those requirements will be satisfied. A video demonstrating the Seldonian Toolkit is available at https://youtu.be/wHR-hDm9jX4/.