UC Davis

Humans and animals can flexibly attribute causes to past choices upon receiving outcomes and use learned information to make predictions in novel scenarios. In this doctoral dissertation, I present evidence on how the brain facilitates these processes by activating a network of regions during feedback, each supporting distinct functions in learning and decision making. Chapter 1 reviews relevant literature on key brain regions and their functions involved in causal learning and memory. Chapter 2 investigates the neural mechanisms in contingent learning, demonstrating parallel reinstatements in the frontal cortex for observed and inferred choices, with subregions in the orbitofrontal cortex representing distinct dimensions of information. It also highlights how the dopaminergic midbrain reflects individual differences in inferential ability. Chapter 3 extends beyond the hierarchical task structure used in Chapter 2 and assesses how contingent learning and abstraction are shaped by hierarchical levels in environments with multiple causal layers. Chapter 4 proposes future research directions based on these findings. Collectively, these results elucidate neural mechanisms in contingent learning, essential for building and updating cognitive maps.