A Neurocomputational Approach to Trained and Transitive Relations in Equivalence Classes

Area of Science:

• Cognitive Science
• Neuroscience
• Behavioral Psychology

Background:

• Stimulus equivalence involves grouping perceptually distinct stimuli that are functionally interchangeable.
• Equivalence relations can be learned directly or derived transitively (e.g., A=B, B=C implies A=C).

Purpose of the Study:

• To propose a neurocomputational model for the fundamental learning mechanisms underlying stimulus equivalence class formation.
• To elucidate how relatedness within equivalence classes develops for both trained and derived stimulus relations.

Main Methods:

• Development of a neurocomputational model to simulate stimulus equivalence.
• Simulation of three classic stimulus equivalence studies, including typical and atypical populations and nodal distance effects.

Main Results:

• The model demonstrates a mechanism for selectively strengthening stimulus associations, even without direct co-presentation.
• It explains the development of relatedness within equivalence classes based on learned and derived relations.

Conclusions:

• The model links stimulus equivalence to Hebbian learning and categorization theories.
• It highlights the utility of modeling stimulus equivalence for understanding learning protocols and their effects.