INDEX
    Explanations

    The neuron fires on HTML markup tokens (especially closing tags like </div>), i.e. it detects the presence of HTML tags.

    New Auto-Interp
    Negative Logits
    ENTRY
    -0.07
     hardly
    -0.07
    hive
    -0.07
    274
    -0.06
     нового
    -0.06
    či
    -0.06
    ุบ
    -0.06
    -0.06
     However
    -0.06
    -0.06
    POSITIVE LOGITS
     Philippine
    0.07
     хвилин
    0.06
    >
    ↵
    ↵
    ↵
    0.06
    Islam
    0.06
     nikdo
    0.06
    نتی
    0.06
    eresa
    0.06
     світі
    0.06
    .Failure
    0.06
    retrieve
    0.06
    Act Density 0.011%

    No Known Activations