Active inference and psychodynamics: a novel integration with applications to depression and stress disorders.

This paper introduces Active Intersubjective Inference (AISI), a novel framework that integrates psychodynamic theory with predictive processing to explain self-identity construction and psychopathology. AISI posits that the self emerges from recursive inferences about how others perceive us (second-order self), interacting bidirectionally with interoceptive processes. We map psychodynamic phenomena (e.g., transference, projection) onto neurocomputational mechanisms and apply AISI to Major Depressive Disorder (MDD) and Post-Traumatic Stress Disorder (PTSD), highlighting distorted second-order inference as a core dysfunction. Therapeutic implications include psychodynamic therapy, cognitive-behavioral approaches, and psychedelic-assisted treatments to enhance inference flexibility. AISI bridges psychodynamic insights with the NIMH Research Domain Criteria, offering a testable model for precision psychiatry and future clinical trials.