INDEX
    Explanations

    The neuron fires on the word “SOFTWARE,” i.e. it detects occurrences of “SOFTWARE” (typically in license disclaimers).

    New Auto-Interp
    Negative Logits
     ده
    -0.07
    amarin
    -0.06
    fortune
    -0.06
     реш
    -0.06
    \Migration
    -0.06
    "},
    ↵
    -0.06
     ----------------------------------------------------------------------------------------------------------------
    -0.06
     مواطنة
    -0.06
     Lumia
    -0.06
    .booking
    -0.05
    POSITIVE LOGITS
     delete
    0.08
    .int
    0.07
     лок
    0.07
     Nicar
    0.07
     optional
    0.07
    Optional
    0.07
     soul
    0.07
     Swap
    0.07
    сом
    0.06
    0.06
    Act Density 0.003%

    No Known Activations