Emotional learning: A computational model of the amygdala

Abstract

Emotional learning is driven by the interaction between the amygdala and the orbitofrontal cortex, operating within a functional two-process framework. The amygdala facilitates the acquisition of excitatory associations between sensory stimuli and their emotional consequences, utilizing rapid, coarse inputs from the thalamus alongside highly analyzed representations from the sensory cortex. Complementing this, the orbitofrontal cortex provides inhibitory control to suppress associations that are no longer valid, enabling extinction when expected reinforcements are omitted. A computational model incorporating these neurophysiological principles successfully replicates behavioral phenomena such as habituation, acquisition, extinction, and blocking. Furthermore, simulated lesions to the prefrontal and sensory cortices highlight their critical roles in discriminative learning and the regulation of emotional responses. By integrating data from learning theory and neurophysiology, this model demonstrates that emotional processing follows systematic computational rules that guide adaptive behavior. – AI-generated abstract.