Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    deep

    np_max-act · gemini-2.0-flash

    This neuron detects Spanish technical terms referring to how deep or how far something penetrates (e.g. “profundidad,” “penetración,” “penetrar”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    χε
    -0.07
    esis
    -0.07
    ंट
    -0.07
     alist
    -0.06
    elt
    -0.06
    outil
    -0.06
    \e
    -0.06
     tik
    -0.06
    販売
    -0.06
     если
    -0.06
    POSITIVE LOGITS
     penetrate
    0.07
     สามารถ
    0.06
    (property
    0.06
    ezpeč
    0.06
    政府
    0.06
     деся
    0.06
     RuntimeMethod
    0.06
     бли
    0.06
    (btn
    0.06
     bilim
    0.06
    Act Density 0.013%

    No Known Activations