The neuron systematically lights up on semantically “heavy” or informative tokens—especially content words like nouns and adjectives (and even punctuation like commas) that carry high information weight.