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

    Neuronpedia

    INDEX
    Explanations

    function and code definitions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     mal
    0.62
     vai
    0.56
    簋
    0.56
     الداخلية
    0.54
     watch
    0.54
     outweigh
    0.53
     stay
    0.53
     serve
    0.53
    Stay
    0.52
     निगरानी
    0.52
    POSITIVE LOGITS
    损失
    0.57
    kowej
    0.56
     सऊदी
    0.56
     BUR
    0.55
    詛
    0.55
    बंध
    0.55
    奭
    0.55
     ダンロップ
    0.54
    Pitch
    0.54
    <0xB7>
    0.54
    Act Density 0.197%

    No Known Activations