Neural mechanisms of learning, planning, and decision-making (360G-Wellcome-220296_Z_20_Z)

This proposal examines the neural mechanisms supporting decision-making and prospective planning. We will examine how prefrontal cortex (PFC), hippocampus, and entorhinal cortex (EC) interact to support these processes. We will examine how non-human primates (NHPs) make choices in large decision spaces, particularly when novel choice-values have to be inferred ‘online’. We will test different models of value-coding, particularly whether PFC uses a ‘place-like’ and ‘grid-like’ code to construct cognitive maps of values spaces. We will examine how NHPs make ‘online’ choices when sequentially navigating between stimuli/states as rewards move or paths blocked. We will test whether ‘replay’ provides a neural mechanism supporting model-based planning. We will use Transcranial Ultrasound Stimulation to selectively disrupt regions of PFC/hippocampus/EC to examine its effect on neural selectivity and behaviour. These tasks are high-dimensional, yet amenable to mathematical description, and will be combined with high-density recordings to map these computations. Exp.3 will integrate our home-cage training system with wireless data-logging to record neural data continuously, across tasks and sleep, to examine how neural signatures change across days with learning, and acquisition of ‘learning set’. This provides the technology to continuously map the NHP brain during performance of diverse and naturalistic tasks, radically transforming primate neuroscience.