INDEX
    Explanations

    instances of the word 'some.'

    New Auto-Interp
    Negative Logits
    swer
    -0.18
    uers
    -0.16
     someone
    -0.16
    umat
    -0.15
    اتÛĮ
    -0.15
     something
    -0.15
    lif
    -0.15
    ãĥªãĥ¼ãĤº
    -0.14
    esi
    -0.14
    _some
    -0.14
    POSITIVE LOGITS
    ones
    0.36
    place
    0.33
    /all
    0.32
    kind
    0.25
     sort
    0.24
    hw
    0.24
    ht
    0.23
    ONES
    0.22
    -times
    0.22
     of
    0.22
    Act Density 0.098%

    No Known Activations