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

    Neuronpedia

    INDEX
    Explanations

    work live parts

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     vetting
    0.56
    德國
    0.55
    Якщо
    0.55
    ську
    0.54
    ဟုတ်
    0.54
    лянчук
    0.54
     vetted
    0.54
     серпня
    0.53
    Deaths
    0.53
     attempting
    0.53
    POSITIVE LOGITS
     optimally
    0.52
     impresses
    0.50
    赋予
    0.47
    借助
    0.47
     impress
    0.47
     compactly
    0.47
     interplay
    0.47
    钥匙
    0.46
     شود
    0.46
     Particularly
    0.45
    Act Density 0.136%

    No Known Activations