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

    Neuronpedia

    INDEX
    Explanations

    dates and separators

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ล
    0.72
    س
    0.67
    ס
    0.66
    s
    0.65
    我
    0.64
    с
    0.62
    8
    0.61
    م
    0.60
    But
    0.59
    ก็
    0.59
    POSITIVE LOGITS
    is
    0.68
    ピオン
    0.57
    퇸
    0.56
     irregularities
    0.56
    umia
    0.55
    ли
    0.54
     fibres
    0.54
    তাহাদের
    0.54
     Writes
    0.54
    ́u
    0.54
    Act Density 0.009%

    No Known Activations