INDEX
    Explanations

    The neuron fires on emphasized or strongly intensifying tokens (words marked or used to add emphasis).

    New Auto-Interp
    Negative Logits
    FON
    0.40
    각형
    0.40
    รายงาน
    0.39
    SampleSize
    0.38
    KeyCode
    0.37
     })}
    0.37
    0.37
    AsyncTask
    0.36
    0.36
     Muitos
    0.36
    POSITIVE LOGITS
     afe
    0.48
     Screw
    0.41
     screw
    0.40
    screw
    0.40
     wen
    0.38
    чки
    0.38
     nha
    0.37
     Couch
    0.37
     apro
    0.37
     ell
    0.37
    Act Density 0.000%

    No Known Activations