INDEX
    Explanations

    occurrences of the word "per" and its variations

    New Auto-Interp
    Negative Logits
    abile
    -0.17
    wright
    -0.15
    .tt
    -0.15
    scal
    -0.15
    egg
    -0.15
    輪
    -0.14
    egend
    -0.14
    iard
    -0.14
    sworth
    -0.14
    bling
    -0.14
    POSITIVE LOGITS
    ohl
    0.21
    anto
    0.18
    quisites
    0.18
    vez
    0.15
     Tet
    0.15
    ceptions
    0.15
    hn
    0.15
    oxide
    0.14
    è©
    0.14
    man
    0.14
    Act Density 0.086%

    No Known Activations