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

    Neuronpedia

    INDEX
    Explanations

    originally followed by past actions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    comme
    1.86
    वर
    1.70
     cherish
    1.69
    HING
    1.68
    戴
    1.67
    дцать
    1.67
    ine
    1.67
    ᇁ
    1.65
     alas
    1.62
    يس
    1.60
    POSITIVE LOGITS
    ÇÕES
    1.74
    운데
    1.67
    ळ
    1.62
    u
    1.62
     uiteindelijk
    1.58
    ronectin
    1.57
     Tät
    1.52
    rotated
    1.52
    Indeed
    1.52
     Indeed
    1.49
    Act Density 0.002%

    No Known Activations