INDEX
    Explanations

    the word "even" and variations of its usage in different contexts

    New Auto-Interp
    Negative Logits
    BBBB
    -0.17
    ivities
    -0.16
    erge
    -0.16
    unker
    -0.15
    senal
    -0.15
    FFFFFFFF
    -0.15
    bersome
    -0.15
    idal
    -0.14
    ivating
    -0.14
    htub
    -0.14
    POSITIVE LOGITS
    -handed
    0.17
    eyh
    0.16
    irth
    0.15
    LS
    0.14
    fois
    0.14
     worse
    0.14
    Criterion
    0.13
    327
    0.13
    eyen
    0.13
    걸
    0.13
    Act Density 0.019%

    No Known Activations