Matthew Gardner

This study uses computational modeling and optogenetics to show that dopamine neuron stimulation promotes learning, supporting the prediction error hypothesis over the rewarding value hypothesis.

This study investigates how hippocampal output affects schema formation in orbitofrontal cortex, finding that inactivation during learning accelerates schema correlates while post-learning inactivation has no effect, supporting a parallel model of cognitive map formation.

The paper investigates the role of the orbitofrontal cortex in creating cognitive maps by showing that its inactivation disrupts specific cue-outcome associations but not model-based learning overall.

This study identifies prospective coding in rat orbitofrontal cortex for predicting future outcomes during an odor sequence task.

This paper shows that orbitofrontal cortex activity during cue preconditioning is necessary for value inference in probe tests, indicating it encodes associative information beyond value.

The paper provides causal evidence that dopamine transients from reward-predictive cues function as prediction errors by showing their role in second-order conditioning.

This study shows that rat orbitofrontal cortex activity represents both a value-invariant task structure and expected value during an odor sequence task.

This paper proposes a new theory of dopamine function that extends prediction error to include sensory predictions, bridging model-based and model-free reinforcement learning.