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

    Neuronpedia

    INDEX
    Explanations

    overtly sexual or negative

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ه
    0.59
    ה
    0.53
    an
    0.52
    secrets
    0.51
    ing
    0.51
    contenedor
    0.51
    grounds
    0.50
    founded
    0.49
     stents
    0.48
    ＞
    0.48
    POSITIVE LOGITS
     방법에
    0.49
     Persia
    0.49
     উপায়
    0.48
     Atomnoj
    0.47
     ವಿಧಾನ
    0.47
    󠁥
    0.46
    राक
    0.45
    olik
    0.45
    邁
    0.45
    átku
    0.45
    Act Density 0.003%

    No Known Activations