In this talk, I will discuss our work into the neurocomputational mechanisms underlying psychotic symptoms of schizophrenia spectrum disorders, including hallucinations, delusions, and insight impairments. I will briefly introduce basic models of perceptual inference and present evidence for inferential prior-related biases in hallucinations. I will then discuss our more recent work on similar inferential biases towards prior beliefs at higher levels of abstraction, which we have empirically linked to delusions and delusion-like ideation. I will emphasize unpublished fMRI work investigating the neural bases of delusion-related prior overweighting, which suggests that variability in the neural representation of prior beliefs relates to prior overweighting across healthy and patient populations and further scales with delusion severity in the latter. In the second part, I will discuss shortcomings of current models of perceptual inference in psychosis, with an emphasis on explanatory gaps for the clinical phenomenon of insight, impairments in which are characteristic of psychotic disorders. I will discuss a novel paradigm and modeling framework we have designed to capture this phenomenon, as well as our future plans for extensions in clinical populations and theory development.


Dr. Guillermo Horga 

Associate Professor of Psychiatry, Department of Psychiatry, Columbia University College of Physicians & Surgeons

Founding Co-director, Clinical Cognitive Computational Neuroscience Center (C3N) at Columbia Psychiatry


Guillermo Horga – Perceiving without believing: inference, psychosis, and insight