INDEX
    Explanations

    the character sequence "thro" inside tokens (a common subword in medical/biological terms).

    New Auto-Interp
    Negative Logits
     Saunders
    -1.10
    asha
    -0.76
    <bos>
    -0.63
    ople
    -0.58
    PerformLayout
    -0.56
     CreateTagHelper
    -0.53
    roll
    -0.51
     possible
    -0.48
    бре
    -0.47
    gms
    -0.46
    POSITIVE LOGITS
     للمعارف
    0.81
    AutoScaleMode
    0.78
    0.63
    mybatisplus
    0.61
     DBNull
    0.60
     Vikipedi
    0.60
    saraba
    0.60
    gonic
    0.59
     كومونز
    0.58
     conformidad
    0.57
    Act Density 0.014%

    No Known Activations