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

    Neuronpedia

    INDEX
    Explanations

    Code

    np_max-act · gemini-2.0-flash

    The neuron responds whenever a token contains or is part of a numeric literal (digits).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     відповідаль
    -0.07
    onis
    -0.06
     insects
    -0.06
     Graz
    -0.06
     hysteria
    -0.06
     Career
    -0.06
    .legend
    -0.06
    unda
    -0.06
     titan
    -0.05
    ogh
    -0.05
    POSITIVE LOGITS
    阵
    0.07
    traction
    0.07
    WithType
    0.07
     Elev
    0.07
    NSError
    0.07
     conditional
    0.07
    NBC
    0.07
     Duch
    0.06
    ฟ
    0.06
     Ni
    0.06
    Act Density 0.055%

    No Known Activations