Waterloo.ai Seminar: Ryan Smith on Emotion-cognition interactions as deep active inference

Friday, January 17, 2020 5:30 am - 5:30 am EST

Please join us for the next institute seminar on Friday, January 17 at 10:30am in DC 1302.

Title: Emotion-cognition interactions as deep active inference

Abstract: Individuals are known to differ in their ability to understand or be aware of their own emotions. Low emotional awareness (EA) is also a known correlate and risk factor for the development of a range of psychiatric disorders and co-morbid medical conditions. According to the “three-process model” of affective processing, low EA may follow from maladaptive functioning in the way affective bodily responses are generated, the way they are subsequently represented/conceptualized, and/or in the way top-down attentional processes are applied to one’s own emotions. Previous neuroimaging work has linked individual differences in EA to structural and functional differences in large-scale neural networks subserving somatovisceral regulation, interoception, conceptualization, and cognitive control. In this talk, I will briefly introduce the three-process model. I will then describe how this model can be understood in computational terms within the active inference framework. Specifically, I describe a deep (active) inference model that reproduces the cognitive-emotional processes and self-report behaviors associated with EA. I will then present simulations to illustrate (seven) distinct mechanisms that (either alone or in combination) can produce phenomena – such as somatic misattribution, coarse-grained emotion conceptualization, and constrained reflective capacity – characteristic of low EA. These simulations suggest that the clinical phenotype of impoverished EA can be reproduced by dissociable computational processes. The possibility that different processes are at work in different individuals suggests that they may benefit from distinct clinical interventions. This may be a useful step toward identifying which processes operate in different individuals – and providing a principled basis for personalized precision medicine.

Speaker Bio:

Ryan Smith is an associate investigator at the Laureate Institute for Brain Research in Tulsa, Oklahoma. He received his Bachelor’s degree in Psychology from Arizona State University in 2010. He subsequently completed three graduate degrees from the University of Arizona between 2011 and 2015. This included a Master’s degree in Neuroscience, a 2nd Master's degree in Philosophy, and a PhD in Psychology.  He subsequently completed a 3-year postdoctoral fellowship in the Psychiatry Department of the University of Arizona College of Medicine. He then spent 4 months at University College London doing a visiting fellowship with Dr. Karl Friston. 

Dr. Smith accepted his current investigator position at the Laureate Institute in February 2019, where he is currently building a lab focused on applying cognitive, computational, and neurophysiological methods to study brain-body interactions and emotion-cognition interactions in mood and anxiety disorders. Some of his major interests involve understanding how the brain generate emotions, how people subsequently recognize and become aware of their own emotions, and how these mechanisms may be effected in emotional disorders. One major aim of his research is to characterize emotion-cognition interactions within computational models and identify means of identifying aberrant computational mechanisms at the single-patient level to inform both treatment selection and treatment development. Currently, Dr. Smith has authored 63 publications within peer-reviewed journals and scholarly books, based in part on the work described above. These include empirical studies as well as theoretical contributions, proposing neurocognitive and computational models of conscious and unconscious emotion, neurovisceral integration, emotional intelligence, and depression.

Date and Time:

Friday, January 17, 2020
10:30 AM - 11:30 AM

Location: DC 1302
Light refreshments will be available.