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

    Neuronpedia

    INDEX
    Explanations

    Math, numbers, and conversions

    np_max-act · gemini-2.0-flash

    The neuron activates on tokens that are parts of numeric literals (digits and decimal-/base-notation), i.e. it detects number tokens.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    adors
    -0.07
     parms
    -0.06
    >N
    -0.06
    EEE
    -0.06
     RESOURCE
    -0.06
     scary
    -0.06
    INARY
    -0.06
     них
    -0.06
     HB
    -0.06
    UILD
    -0.06
    POSITIVE LOGITS
    _visit
    0.07
     branding
    0.07
     phil
    0.06
    上
    0.06
     ott
    0.06
    emplace
    0.06
     повыш
    0.06
     прест
    0.06
    [])
    0.06
     aktu
    0.06
    Act Density 0.024%

    No Known Activations