INDEX
    Explanations

    early church

    This neuron activates on section and chapter headings or other formal document titles (e.g. book names, chapter numbers, study-week labels).

    New Auto-Interp
    Negative Logits
     Стар
    -0.06
    COMMAND
    -0.06
    eten
    -0.06
    -0.06
     INITIAL
    -0.06
    -0.06
    >(_
    -0.06
    predicted
    -0.06
    руют
    -0.06
     نسب
    -0.06
    POSITIVE LOGITS
    Chinese
    0.07
     margins
    0.06
    ίγ
    0.06
    ogue
    0.06
     бли
    0.06
     Chinese
    0.06
     nog
    0.06
    ��
    0.06
     polož
    0.06
    _inst
    0.06
    Act Density 0.013%

    No Known Activations