The ancient oracle Sibylla wrote her prophecies on leaves. Scattered by the wind, Sibylla's prophecies were difficult to read and often impossible to piece together—however, their accuracy remained intact, whether understood or not. Similarly, the cerebellum generates accurate predictions about the world, without us being aware of how it does it. The SIBYLLA project aims to shed light on the mechanisms underlying context-based predictive processing in the cerebellum, using ultra-high resolution structural and functional magnetic resonance imaging (MRI) at 7 Tesla. The project has two main goals: (1) to map cerebellar regions specifically involved in different types of implicit learning (i.e., identifying general patterns of events versus social-specific contextual learning), and (2) to establish functional-structural relationships in predictive learning networks in the human brain.