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

    Neuronpedia

    INDEX
    Explanations

    to

    np_max-act · gemini-2.0-flash

    The neuron activates on numeric tokens—digits or numeral-related sequences (like “1,” “2.1,” or version numbers).

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

    document-structure and technical tokens such as section/chapter numbers, timestamps, code/config markers, and other layout or formatting elements.

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
     přech
    -0.07
     systém
    -0.07
    �
    -0.07
    omorphic
    -0.06
    atsby
    -0.06
    ヨ
    -0.06
     nedeniyle
    -0.06
    _CLASS
    -0.06
    ظف
    -0.06
    ่างก
    -0.06
    POSITIVE LOGITS
     farewell
    0.07
     Shar
    0.07
     Legislation
    0.07
    .ReadUInt
    0.06
    Shar
    0.06
     exporter
    0.06
    اشین
    0.06
     Gould
    0.06
     Куб
    0.06
     generals
    0.06
    Act Density 0.801%

    No Known Activations