Proceedings of the Annual Meeting of the Cognitive Science Society

The ability to associate words is an important cognitive skill.In this study we investigate different methods for representingword associations in the brain, using the Remote AssociatesTest (RAT) as a task. We explore representations derived fromfree association norms and statistical n-gram data. Althoughn-gram representations yield better performance on the test, acloser match with the human performance is obtained with rep-resentations derived from free associations. We propose thatword association strengths derived from free associations playan important role in the process of RAT solving. Furthermore,we show that this model can be implemented in spiking neu-rons, and estimate the number of biologically realistic neuronsthat would suffice for an accurate representation.