The neuron strongly activates on standalone technical‐domain nouns (like “medium,” “effect,” “type,” “voltage,” “semiconductor,” “areas,” “power,” “imaging,” etc.) that serve as key concepts or headings in scientific and patent‐style texts.
The neuron strongly activates on key technical nouns (e.g. “medium,” “effect,” “potential,” “power,” “imaging,” “type”) that label major scientific concepts or components in a document.
