INDEX
    Explanations

    The neuron responds to evaluative adjectives and adverbs that emphasize or intensify positive qualities (e.g. “inspiring,” “truly,” “refine”).

    New Auto-Interp
    Negative Logits
     Điều
    -0.07
     таком
    -0.06
     Goddess
    -0.06
    -0.06
     спортив
    -0.06
    ับม
    -0.06
     Retry
    -0.06
     tide
    -0.06
     eens
    -0.06
     gras
    -0.06
    POSITIVE LOGITS
    +"/"+
    0.06
    [^
    0.06
     vk
    0.06
    AAAAAAAA
    0.06
     pours
    0.05
    .enterprise
    0.05
     PSP
    0.05
    antd
    0.05
     ThemeData
    0.05
    (rel
    0.05
    Act Density 1.096%

    No Known Activations