CBBM Lecture - Controllability shapes human learning and decision-making under uncertainty

by Valentin Wyart, Inserm/ École Normale Supérieure Paris

 will take place on Tuesday, April 19th, 2022, Time: 3 pm German time (15:00 hours)

Host: Jonas Obleser
Department of Psychology

Abstract: Making accurate decisions requires identifying the cause of sensory observations, but also anticipating the consequences of possible actions. Although both cognitive processes can be formalized as hidden-state inference, they are studied in the lab using different experimental frameworks, which makes their comparison difficult. A constitutive, yet rarely considered difference between these two forms of inference lies in the degree of control over information sampling conferred to the decision-maker. In this talk, I will present a series of new findings obtained in my group using an experimental framework where inference with and without control can be compared in otherwise tightly matched conditions. First, I will show that humans perceive the same environment as more stable when inferring its hidden state by interaction with uncertain outcomes (with control) than by observation of equally uncertain cues (without control). Magnetoencephalographic (MEG) activity reflects this cognitive effect in the consistency between sampled information and the inferred hidden state, a neural signal originating from the temporal lobe. I will then compare response switches made with and without control to dissociate information seeking from confounding factors associated with exploration-exploitation dilemmas. Last, I will explain why controllability may be particularly relevant for understanding obsessive-compulsive disorder (OCD), a psychiatric condition often described as a general ‘disease of uncertainty’. Indeed, severe OCD patients show deficits of inference only in the condition with control, where uncertainty concerns the outcomes of their actions. Taken together, these findings indicate that controllability constitutes an important factor for understanding human learning and decision-making under uncertainty.

Speaker information: Valentin Wyart studied cognitive neuroscience at École Normale Supériere, Paris. After his PhD at Université Pierre et Marie Curie, Paris (with Catherine Tallon-Baudry and Stanislas Dehaene) he was a postdoctoral fellow with Christopher Summerfield and Kai Nobre at the University of Oxford. Now he is a Principal Investigator at the Cognitive Computational Neuroscience Lab of the École Normale Supérieure in Paris, France. His research group studies human learning and decision-making under uncertainty, with a particular focus on the sources of cognitive variability both within and between individuals. The research carried out in his group typically combines behavioral modeling with multimodal functional neuroimaging.

Weiss, A., Chambon, V., Lee, J.K., Drugowitsch, J., Wyart, V. Interacting with volatile environments stabilizes hidden-state inference and its brain signatures. Nature Comm. 12, 2228 (2021). https://doi.org/10.1038/s41467-021-22396-6
Rouault, M., Weiss, A., Lee, J.K., Drugowitsch, J., Chambon, V., Wyart, V. Controllability reveals defining features of information seeking. bioRxiv preprint (2021). https://doi.org/10.1101/2021.01.04.425114

Valentin Wyart