INDEX
    Explanations

    This neuron detects occurrences of the phrase “In contrast” signaling a shift or comparison.

    New Auto-Interp
    Negative Logits
    (){
    ↵
    -0.07
     ambos
    -0.07
    .cpp
    -0.06
     foregoing
    -0.06
     albums
    -0.06
    273
    -0.06
    LayoutParams
    -0.06
    /settingsdialog
    -0.06
    _individual
    -0.06
     слух
    -0.06
    POSITIVE LOGITS
     человека
    0.07
    Moment
    0.07
    blast
    0.06
     vaccines
    0.06
     Вот
    0.06
    _style
    0.06
     Putin
    0.06
    spiracy
    0.06
     Sniper
    0.06
    0.06
    Act Density 0.014%

    No Known Activations