PROJECT 1 – Latent cause inference as a fundamental cognitive process Latent cause inference is a fundamental cognitive process that allows us to group experiences together, so that what happened in past events can easily influence what we do in new (similar) situations. Different individuals might group events differently into latent causes, depending on the parameters of latent cause inference in their brain: how readily they create new latent causes to explain novel experiences, how persistent they believe latent causes are across time, and how similar they expect events to be that are ascribed to a shared latent cause. Because latent cause inference is at the heart of perception, learning, evaluation, and action selection, alterations in the process due to extreme parameters may lead to psychopathology. For instance, over-splitting of latent causes can prevent updating of previous knowledge with new contradicting information (see Project 4), possibly leading to a disorder of compulsion (Project 2) or anxiety (Project 3). Parameters of the latent cause inference process may therefore provide useful biomarkers for vulnerability to psychopathology, and this cognitive process may present an important transdiagnostic dimension that should be added to the NIMH Research Domain Criteria (RDoC) framework. In Project 1, our goal is to characterize fundamental individual differences in latent cause inference and relate them to mental health symptom dimensions in a large general population sample (Aim 1.1), test their potential relevance to several key mental health conditions (generalized anxiety, obsessive compulsive disorder, and schizophrenia; Aim 1.2), and understand the circuitry that realizes latent cause inference in the brain, and its alteration in psychiatric conditions (Aim 1.3). Towards these aims, we will use three behavioral tasks, one laboratory task designed specifically to measure latent cause inference (the Microbes Task), and two naturalistic event segmentation tasks that measure individual differences in perception of boundaries in naturalistic streams of events (movies or stories). By characterizing latent cause inference across mental health conditions and using multiple tasks, the results of Project 1 will provide a comprehensive assessment of the potential use of latent cause inference as an RDoC construct. Moreover, data from this project will provide strong constraints for the computational model of latent cause inference developed by Core C, which will be used to analyze data from all projects in this Center. In this way, association of latent cause inference parameters with a variety of symptom dimensions will help constrain and interpret results in Projects 2 and 3, and the results of Aim 1.3 will potentially inform neural recordings in Project 4.