Abstract:It is shown that emotions can usefully be considered as states produced by reinforcing stimuli. The ways in which a wide variety of emotions can be produced, and the functions of emotion, are considered. There is evidence that the amygdala is involved in the formation of stimulus-reinforcement associations, and the orbitofrontal cortex with correcting behavioural responses when these are no longer appropriate because previous reinforcement contingencies change. This evidence comes from the effects of damage to these structures, and from recording the activity of single neurons in these structures in the monkey during the formation and disconnection of stimulus-reinforcement associations. In so far as emotions can be defined as states produced by reinforcing stimuli, then the amygdala and orbitofrontal cortex are seen to be of great importance for emotions, in that they are involved respectively in the elicitation of learned emotional responses, and in the correction or adjustment of these emotional responses as the reinforcing value of environmental stimuli alters. One of the theses advanced is that the changes in emotional behaviour produced by damage to the brain can be analysed and understood by considering how different parts of it function in reinforcement and in the formation and disconnection of stimulus-reinforcement connections. Another thesis is that there is a population of neurons in the amygdala and parts of the temporal lobe visual cortex specialised to respond to faces, and that these neurons may be involved in social and emotional resposes to faces. Some of the outputs of the amygdala and orbitofrontal cortex are directed to the hypothalamus, which not only provides one route for these reinforcing environmental events to produce autonomic responses, but also is implicated in the utilisation of such stimuli in motivational responses, such as feeding and drinking, and in emotional behaviour. Other outputs of the amygdala and orbitofrontal cortex which may enable them to influence behaviour are directed to the striatum, and also back towards some of the cortical regions from which they receive inputs. It is suggested that these latter projections are important in the effects which mood states have on cognitive processing.

A Temporal Difference Reinforcement Learning Theory of Emotion: unifying emotion, cognition and adaptive behavior

A reinforcement learning model of joy, distress, hope and fear

Emotion in reinforcement learning agents and robots: a survey

Human-Robot Emotional Interaction Model Based on Reinforcement Learning

An Emotion-Based Approach to Reinforcement Learning Reward Design

A Theory of Emotion, Its Functions, and Its Adaptive Value

A Theory of Emotion, and its Application to Understanding the Neural Basis of Emotion

Play with Emotion: Affect-Driven Reinforcement Learning

Dissociable neural signals for reward and emotion prediction errors

A generic self-learning emotional framework for machines

On The Brain And Emotion

Expectation to rewards modulates learning emotional words: Evidence from a hierarchical Bayesian model

Emotional Theory of Rationality

Reinforcement Learning with Brain-Inspired Modulation can Improve Adaptation to Environmental Changes

Emotions as abstract evaluation criteria in biological and artificial intelligences

What Are Emotional States, and Why Do We Have Them?

Temporal Difference Models: Model-Free Deep RL for Model-Based Control

Hidden Reward: Affect and Its Prediction Errors as Windows Into Subjective Value

Neural Connectivity Underlying Reward and Emotion-Related Processing: Evidence From a Large-Scale Network Analysis

Reinforcement Learning as a Process-Based Explanation of Symptom Change

A Neurobiological Approach to Emotional Intelligence