INDEX
    Explanations

    unraveling mysteries and clues

    New Auto-Interp
    Negative Logits
    Config
    0.92
    inline
    0.91
    softmax
    0.91
     Breite
    0.90
    0.89
    🔈
    0.89
    0.89
     ширина
    0.88
    字段
    0.86
    }_{
    0.86
    POSITIVE LOGITS
     mysterious
    1.88
     venge
    1.84
     enigmatic
    1.70
     sinister
    1.66
     haunted
    1.62
     mysteriously
    1.62
     flashbacks
    1.61
     supernatural
    1.60
     mysteries
    1.60
     blackmail
    1.59
    Act Density 0.828%

    No Known Activations