Real world decision problems often involve large state spaces, within which we must plan our actions while accounting for the presence of other agents that may help or hurt us. The complexity of these problems, and their solutions, is not always captured in standard decision-making tasks, despite likely being crucial for understanding the processes underpinning mental health problems. I will present work using virtual environments to understand how humans seek reward while avoiding danger posed by other agents. The findings demonstrate that humans use model-based planning to both predict the other agents’ behaviour and determine their own course of action. I will also present preliminary findings linking these complex interactive decision-making processes to transdiagnostic dimensions of mental health.

Senior Research Fellow
Institute of Psychiatry, Psychology & Neuroscience, King’s College London

Toby Wise – Capturing the complexity of interactive decision problems: mechanisms and implications for mental health