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

    Neuronpedia

    INDEX
    Explanations

    Code special characters

    np_max-act · gemini-2.0-flash

    This neuron fires on tokens containing uppercase letters or digits (e.g., version numbers, XML/HTML syntax markers, and all-caps identifiers).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    Equals
    -0.07
    анных
    -0.06
    κη
    -0.06
     Erl
    -0.06
    OME
    -0.06
    램
    -0.06
    رت
    -0.06
     hides
    -0.06
    545
    -0.06
    .ย
    -0.06
    POSITIVE LOGITS
    /dr
    0.07
     bölge
    0.07
     nghiên
    0.06
     Кар
    0.06
     Heidi
    0.06
     durumlarda
    0.06
     hospital
    0.06
     комплекс
    0.06
     směrem
    0.06
     gün
    0.06
    Act Density 0.139%

    No Known Activations