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

    Neuronpedia

    INDEX
    Explanations

    very / quite

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     zodat
    0.73
     допомогти
    0.65
     ताकि
    0.63
     ώστε
    0.63
     نتیجے
    0.62
     żeby
    0.61
    ępnie
    0.60
     然後
    0.59
    簡単な
    0.59
    自分の
    0.57
    POSITIVE LOGITS
     oldukça
    1.12
     काफी
    1.10
     sangat
    1.09
     весьма
    1.03
     rất
    1.01
     quite
    0.96
     очень
    0.96
     довольно
    0.92
     매우
    0.91
    雖
    0.91
    Act Density 0.045%

    No Known Activations