INDEX
    Explanations

    Broad range of topics

    The neuron detects enumerative or example-introducing phrases—words like “including,” “such as,” “tasks,” “services,” or “issues” that signal lists or examples.

    New Auto-Interp
    Negative Logits
    όν
    -0.07
     engr
    -0.07
    061
    -0.07
    Distribution
    -0.06
    -0.06
     konu
    -0.06
     oci
    -0.06
    ρά
    -0.06
    大量
    -0.06
     customs
    -0.06
    POSITIVE LOGITS
     demek
    0.07
     like
    0.06
    Snackbar
    0.06
    /stdc
    0.06
     siyaset
    0.06
     ASM
    0.06
    =mysqli
    0.06
    .functions
    0.06
     sklad
    0.06
    /s
    0.06
    Act Density 0.120%

    No Known Activations