The neuron strongly activates on standalone quoted tokens and other high-salience words or phrases (e.g. protected‐class terms, quoted speech markers, proper names), essentially flagging the boundaries and content of important quoted or named entities.
