INDEX
    Explanations

    questions and expressions of disbelief or frustration

    New Auto-Interp
    Negative Logits
    .nlm
    -0.15
    zent
    -0.15
    Crud
    -0.14
     åĵ
    -0.14
    OPS
    -0.14
    ictory
    -0.14
     Beard
    -0.14
    undry
    -0.14
    setSize
    -0.14
    koli
    -0.14
    POSITIVE LOGITS
     why
    0.34
     Why
    0.26
     how
    0.24
    why
    0.24
     WHY
    0.21
     shouldn
    0.20
     seriously
    0.20
     surely
    0.20
     How
    0.20
    Why
    0.19
    Act Density 0.288%

    No Known Activations