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

    Neuronpedia

    INDEX
    Explanations

    carry

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Jika
    0.78
     있으며
    0.78
     אך
    0.70
    했으며
    0.69
     yalnızca
    0.69
    を含む
    0.67
     nicknamed
    0.66
     adlı
    0.64
     jika
    0.64
     ("
    0.63
    POSITIVE LOGITS
     굉장히
    0.99
    比如说
    0.92
    非常的
    0.89
     sort
    0.86
     really
    0.85
    這樣子
    0.85
     ähm
    0.82
    っていう
    0.80
     somebody
    0.79
     anybody
    0.78
    Act Density 0.002%

    No Known Activations