A computational model reveals classical conditioning mechanisms underlying visual signal detection in rats

Area of Science:

• Cognitive Neuroscience
• Computational Neuroscience
• Animal Behavior

Background:

• Sustained attention is crucial for complex tasks.
• Classical conditioning models offer insights into attention and decision-making.
• Previous models did not fully integrate occasion setting with signal detection.

Purpose of the Study:

• To apply a neural network model of classical conditioning to visual signal detection and discrimination tasks in rats.
• To test the role of visual signals as occasion setters in attention tasks.
• To explain rat behavior in sustained attention paradigms.

Main Methods:

• Utilized a neural network model of classical conditioning (Schmajuk et al., 1996).
• Applied the model to visual signal detection and discrimination tasks.
• Interpreted rat behavior using a sign-tracking approach, focusing on lever-approach responses.

Main Results:

• The model accurately predicted increased hits with signal intensity and decreased hits with trial rate.
• Model performance matched observed lower accuracy in discrimination versus detection tasks.
• Rat behavior confirmed that visual signals act as occasion setters, not direct reward cues.

Conclusions:

• The neural network model successfully describes sustained attention in rats.
• Rats learn to associate specific locations (levers) with expected rewards.
• Visual stimuli function as occasion setters, influencing response selection rather than becoming direct reward associations.