INDEX
    Explanations

    The neuron fires on occurrences of the word “tangent” (or forms like “tangential”).

    New Auto-Interp
    Negative Logits
     Mozart
    -0.07
     امید
    -0.07
    769
    -0.07
     Fever
    -0.07
     Lew
    -0.07
     Schro
    -0.07
     Live
    -0.07
    Crud
    -0.07
     فرد
    -0.06
     bowel
    -0.06
    POSITIVE LOGITS
     Tang
    0.14
     tang
    0.13
     Tan
    0.11
     tangent
    0.10
    Tan
    0.10
     tan
    0.10
    int
    0.08
     Fang
    0.08
    Ang
    0.08
    atter
    0.07
    Act Density 0.007%

    No Known Activations