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

    Neuronpedia

    INDEX
    Explanations

    maintain/keep/impact

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    aná
    1.06
     Поскольку
    0.90
     compleja
    0.88
    ственный
    0.86
    uity
    0.86
    imilar
    0.85
    чное
    0.84
     elétrica
    0.83
     ganancias
    0.83
    iou
    0.83
    POSITIVE LOGITS
    LAND
    0.99
    الصفحه
    0.91
    ش
    0.89
    Strings
    0.88
    Move
    0.87
    ستوى
    0.87
     CHAS
    0.86
    يا
    0.84
    selves
    0.83
    د
    0.83
    Act Density 0.001%

    No Known Activations