Filip's background is in engineering (signal processing) and computer science. Broadly speaking, he is interested in how humans and machines learn and make decisions, and how they ought to do it. During his PhD, he investigated how machines can learn about mental states (in particular, mind wandering) using brain-computer interfaces. Currently, he's investigating how we can use machines to model how people learn from unpleasant experiences, with the goal of better understanding emotions of fear and anxiety, and the mental disorders related to these emotions. To pursue these questions, he employ's psychological experiments (in the lab and online), physiological measurements (e.g., EEG, heart rate, pupil size), and computational modeling (e.g., Bayesian and reinforcement learning models).

In parallel, he has have cultivated an interest in the question of how scientists ought to optimally learn and make decisions. He has contributed statistical methods for evaluating brain-computer interfaces, and methods for algorithmic optimization of psychological experiments. Currently, he is also pursuing questions about the role of automation in psychology (i.e., psychoinformatics), and questions about optimal approaches to scientific inquiry (i.e., meta-science).