Gerit Pfuhl lab

In my lab we study how information is represented in the brain, and why behaviour (sometimes) is suboptimal. In my PhD I have addressed this question across species (including ravens and jackdaws), and now we look within humans at it across the autism-psychosis spectrum by using computational models, behavioural tasks and physiological techniques (pupillometry, heart rate variability). Specifically, we try to identify what contributes to deliberate reasoning and rational behaviour, i.e. understanding motivation, cognitive effort, and attributing uncertainties appropriately. Our starting point is the brain being Bayesian. In a nutshell, brains have a model or models of the world, incoming sensory information is compared to the expectations from the model, and deviations or prediction errors are encoded. Such a discrepancy, i.e. the prediction error, can lead to either updating of the model, or acting to make the information fitting (I see / hear what I want to see / hear). Notably, costs and benefits of acting (or not acting) can be incorporated and might explain motivational differences and hence a window into the ultimate mechanisms of behavior.

Our research can also be applied to avalanche decision-making, here the uncertainties change during the backcountry skiing (volatile environment) and regular updating of the avalanche risk is necessary but often in conflict with the desire to just enjoy the skiing. A fascinating (though sometimes fatal) case of motivated reasoning. You'll find more about our research on adaptive rationality on the CARE homepage.