This neuron fires most strongly on discourse-structuring cue words—tokens like “example,” “what,” “how,” “whether,” “or,” etc.—that introduce an illustration, question, or contrast.