Value generalization in human avoidance learning.

Area of Science:

• Neuroscience
• Cognitive Science
• Psychology

Background:

• Generalization in aversive decision-making helps avoid threats but over-generalization is linked to psychological disorders.
• Understanding the mechanisms of avoidance generalization is crucial for addressing maladaptive behaviors.

Purpose of the Study:

• To dissect the contributions to the generalization of instrumental avoidance using reinforcement learning modeling.
• To investigate the neural circuits and individual differences associated with aversive and neutral feedback generalization.

Main Methods:

• Employed reinforcement learning modeling in two groups of human volunteers (N=26, N=482).
• Analyzed generalization of instrumental avoidance based on aversive and neutral feedback.
• Correlated generalization patterns with self-reported anxiety and intrusive thoughts.

Main Results:

• Avoidance generalization comprises perceptual and value-based processes.
• Value-based generalization further divides into aversive and neutral feedback components.
• Specific brain circuits (sensory cortex, insula, amygdala, vmPFC) are involved.
• Generalization from aversive, but not neutral, feedback is linked to anxiety and intrusive thoughts.

Conclusions:

• Distinct mechanisms mediate generalization in avoidance learning.
• Individual differences in these mechanisms contribute to anxiety.
• This research provides insights into the neurocognitive basis of anxiety disorders.