On the reciprocal interaction between believing and feeling: an adaptive agent modelling perspective.

Area of Science:

• Cognitive Science
• Neuroscience
• Artificial Intelligence

Background:

• Beliefs are typically influenced by cognitive factors like observation and reasoning.
• Affective factors, such as emotions and feelings, also play a significant role in belief formation.
• Neurological theories highlight the interplay between emotions and cognitive processes.

Purpose of the Study:

• To introduce an agent model integrating cognitive and affective factors in belief.
• To explore how emotional responses affect the strength of an agent's beliefs.
• To model the emergence and strengthening of affective influences on beliefs over time.

Main Methods:

• Development of a computational model incorporating two feedback loops: one for emotional processing and another for belief-feeling interaction.
• Utilizing neurological theories on emotion and feeling, including recursive body loops for emotional responses.
• Application of Hebbian learning to model the development of connections between feelings and beliefs.

Main Results:

• The model demonstrates how belief strength is modulated by both informational input and emotional responses.
• Specific agent characteristics can lead to the emergence of optimistic or pessimistic worldviews.
• Affective influences on beliefs can intensify over time through accumulated experiences, mimicking 'gut feelings'.

Conclusions:

• Emotions and feelings are integral to belief formation and strength, not merely secondary effects.
• The proposed model provides a framework for understanding how affective states shape cognitive beliefs.
• This research offers insights into the mechanisms underlying experiential learning and intuitive judgment in agents.