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

    Neuronpedia

    INDEX
    Explanations

    =

    np_max-act · gemini-2.0-flash

    The neuron is picking out numeric literals (especially version numbers and other numbers in assignments) in the manifest.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Blond
    -0.07
    _capabilities
    -0.07
     Cos
    -0.07
     ecl
    -0.07
    Cla
    -0.06
     French
    -0.06
     khí
    -0.06
     audience
    -0.06
    ůž
    -0.06
    ジョ
    -0.06
    POSITIVE LOGITS
    сер
    0.06
    empor
    0.06
    [][
    0.06
     ~(
    0.06
     CPU
    0.06
    guest
    0.06
     zwar
    0.06
    empt
    0.06
    inoa
    0.06
     SYSTEM
    0.06
    Act Density 0.013%

    No Known Activations