INDEX
    Explanations

    comparison of "never larger"

    New Auto-Interp
    Negative Logits
    ات
    1.63
    ة
    1.63
    ная
    1.53
    ir
    1.50
    1.48
    >**
    1.47
    س
    1.47
    1.44
    сны
    1.43
    1.43
    POSITIVE LOGITS
    C
    1.95
     an
    1.84
    O
    1.73
     a
    1.71
     been
    1.71
     Kabhi
    1.70
    S
    1.66
    D
    1.64
    I
    1.61
    U
    1.59
    Act Density 0.647%

    No Known Activations