Emotion and allostatic control: An active inference account of emotion regulation

Emotions are a central aspect of human experience and they serve as inherently regulatory processes that play a crucial role in supporting allostasis and the adaptive regulation of bodily and physiological states. Here, we propose a novel comprehensive account of emotion regulation - encompassing both implicit and explicit processes - and its role in allostatic control, within the active inference framework. First, we discuss empirical evidence highlighting the roles of emotion regulation in supporting allostasis and the preservation of organismal integrity. Second, we introduce active inference and discuss how emotions have been characterized within it. We emphasize that within active inference, emotions are closely tied to allostatic control, serving as multimodal inferential strategies that support adaptive responses to environmental stressors while preserving biological integrity - through the minimization of interoceptive, proprioceptive and exteroceptive prediction errors, or more formally free energy. Third, we present a novel account of emotion regulation through active inference, that links implicit and explicit regulation to two non-mutually exclusive types of free energy minimization - retrospective/variational free energy and prospective/expected free energy - that underpin distinct regulatory processes. In addition, our proposal systematically maps the core strategies of Gross's process model of emotion regulation to active inference constructs, while also extending this model to incorporate missing elements. Finally, we briefly discuss how linking emotion regulation to active inference offers a novel perspective on maladaptive emotional behavior in psychopathology.