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

    Neuronpedia

    INDEX
    Explanations

    do

    np_max-act · gemini-2.0-flash

    The neuron fires on words that denote or introduce actions or procedures (e.g. “do,” “method,” “adding,” “comprising,” “use,” “invention”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    sentences or phrases that introduce or describe an invention or formal technical matter (patent/boilerplate intro language).

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
     Farr
    -0.07
    Leading
    -0.07
     Barth
    -0.06
    Nombre
    -0.06
     Osw
    -0.06
    他們
    -0.06
    _shapes
    -0.06
     JM
    -0.06
    Av
    -0.06
     WL
    -0.06
    POSITIVE LOGITS
    entar
    0.07
    [↵
    0.07
    查看摘要
    0.06
     oversee
    0.06
    لیه
    0.06
    .getTime
    0.06
    [T
    0.06
    .ToolStrip
    0.06
    쪽
    0.06
    (push
    0.06
    Act Density 0.278%

    No Known Activations