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

    Neuronpedia

    INDEX
    Explanations

    code constructs

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     міся
    0.71
    اد
    0.69
    ря
    0.68
    вої
    0.67
    влека
    0.67
     адміністра
    0.66
     швид
    0.66
     музи
    0.66
     радо
    0.66
     παιδιά
    0.66
    POSITIVE LOGITS
    is
    0.67
     postdoctoral
    0.67
     von
    0.67
    MA
    0.64
     nanom
    0.64
    el
    0.64
     seu
    0.64
     seinen
    0.62
     word
    0.62
     dari
    0.61
    Act Density 0.000%

    No Known Activations

    This feature has no known activations.