INDEX
    Explanations
    New Auto-Interp
    Negative Logits
    лиÑĪком
    -0.16
    ond
    -0.15
    edy
    -0.15
    verse
    -0.14
     Roths
    -0.14
    ilion
    -0.14
    ÑĢем
    -0.13
    ÑĢд
    -0.13
    gi
    -0.13
     Ha
    -0.13
    POSITIVE LOGITS
    QL
    0.16
    amat
    0.15
     front
    0.13
    彩
    0.13
    spr
    0.13
    Ñĩик
    0.13
    aisal
    0.13
     necessary
    0.13
    rame
    0.13
    ipc
    0.12
    Act Density 0.099%

    No Known Activations