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

    Neuronpedia

    INDEX
    Explanations

    articles

    np_max-act · gemini-2.0-flash

    This neuron detects nouns referring to social groups/organizations or technical subject labels (e.g., "Native", "tribal", "Indians", "programming", "language").

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
    Từ
    -0.07
    	or
    -0.06
    =Y
    -0.06
     candy
    -0.06
    สร
    -0.06
     epoxy
    -0.06
    plate
    -0.06
    =".$_
    -0.06
     Leading
    -0.06
     التف
    -0.06
    POSITIVE LOGITS
     MemoryStream
    0.07
     Lancaster
    0.07
     آمریک
    0.06
     nederland
    0.06
    .grid
    0.06
    ActiveSheet
    0.06
     dout
    0.06
    ělí
    0.06
    _SUPPORT
    0.06
    (parseInt
    0.06
    Act Density 0.804%

    No Known Activations