The paper that this article is referring to looks interesting, they performed single neuron recordings in neurosurgical patients. But I am wondering if these findings scale to people with large working memory capacities.