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

    Neuronpedia

    INDEX
    Explanations

    image

    np_max-act · gemini-2.0-flash

    The neuron strongly activates on numeric tokens (especially version numbers and other digit-only or digit-heavy values).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    ابت
    -0.06
     který
    -0.06
    ObjectOfType
    -0.06
    klär
    -0.06
    /dev
    -0.06
    Platform
    -0.06
     Rookie
    -0.06
     dbContext
    -0.05
    ndern
    -0.05
    thank
    -0.05
    POSITIVE LOGITS
     ПО
    0.06
    +
    0.06
    ,L
    0.06
     －
    0.06
    }${
    0.06
    lâm
    0.06
     notifying
    0.06
    Michelle
    0.06
    чої
    0.06
    Including
    0.06
    Act Density 0.002%

    No Known Activations