Towards an understanding of optimism bias: a computational approach

This thesis investigates how optimism bias, which improves health and well-being by promoting engagement in the world, can be an intervention for depression and anxiety using methods from computational psychiatry. I use the active inference framework to demonstrate how we can model and simulate optimism bias in different task settings, providing the groundwork for further research in clinical settings. Using both empirical and interdisciplinary approaches, I show how optimism bias is a promising treatment target.