Prisoner of the present: Borderline personality and a tendency to overweight cues during Bayesian inference.

Recent work has examined the computational mechanisms underlying borderline personality (BP). However, this research has been confined to specific tasks. A computational analysis of BP's mental processes as they broadly unfold in everyday life is lacking. Here, a computational model of BP is proposed, which describes patients' everyday-life mental experience at large. Grounded on Bayesian inference, the proposal is that BP sufferers attribute excessive weight to cues considered to infer life contexts (e.g., to infer whether a cooperation or competition context is ongoing). Remarkably, model simulations demonstrate that this idea accounts for several characteristics of BP, from extreme oscillations in identity, affect, and behavior, to dysfunctional interpersonal cycles. Altogether, the article offers a framework to interpret the broad, everyday life computational mechanisms underlying BP. This can inspire theoretical and empirical research and can help understanding how clinical interventions for BP work, thus contributing to refine such interventions. (PsycInfo Database Record (c) 2022 APA, all rights reserved).