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

    Neuronpedia

    INDEX
    Explanations

    edge cases

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    有可能
    1.79
     écrit
    1.77
     éclat
    1.76
    不太
    1.74
     اين
    1.73
    จะ
    1.67
     inputStream
    1.66
     いや
    1.66
     действие
    1.65
    }}}{\
    1.64
    POSITIVE LOGITS
    ра
    2.34
    و
    2.31
    ut
    2.13
    é
    2.11
    r
    1.97
    きた
    1.92
    嵋
    1.90
    ES
    1.78
    ർ
    1.77
    ен
    1.77
    Act Density 0.736%

    No Known Activations