INDEX
    Explanations

    The neuron is triggered by tokens that look like file names or URLs—i.e. code or resource paths with familiar extensions (e.g. “.cc”, “.jar”, “.png”) or domain‐style strings.

    New Auto-Interp
    Negative Logits
    -0.07
    ведите
    -0.07
    finite
    -0.06
     ذكر
    -0.06
     rospy
    -0.06
    URRED
    -0.06
     thru
    -0.06
    RT
    -0.06
     FK
    -0.06
    _aliases
    -0.06
    POSITIVE LOGITS
    وال
    0.06
    اورزی
    0.06
    ')?>
    0.06
    _Code
    0.06
    мос
    0.06
     подраз
    0.06
     olmuş
    0.06
     đứ
    0.06
     BX
    0.06
    ковод
    0.06
    Act Density 0.033%

    No Known Activations