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

    Neuronpedia

    INDEX
    Explanations

    always, maintain, eliminate

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     nouns
    0.82
     sevent
    0.77
    nouns
    0.76
     symbolic
    0.76
     compra
    0.75
     life
    0.75
     noun
    0.74
     stort
    0.72
    ো
    0.72
     जीवन
    0.71
    POSITIVE LOGITS
     नेहमी
    0.83
    常に
    0.82
    破壊
    0.82
     ہمیشہ
    0.80
     удержи
    0.78
     eliminating
    0.75
    Elim
    0.74
    合わせた
    0.73
    总是
    0.73
    维持
    0.72
    Act Density 0.002%

    No Known Activations