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

    Neuronpedia

    INDEX
    Explanations

    n1 = (i - 1) * nx + j

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ש
    2.17
    ла
    2.03
    ли
    1.93
    ا
    1.74
    aa
    1.66
    tı
    1.59
    یر
    1.52
    ast
    1.51
    ae
    1.51
    ore
    1.44
    POSITIVE LOGITS
    IN
    1.19
     ومع
    1.18
     irrepar
    1.16
    ER
    1.15
    ৫
    1.13
    泗
    1.11
    EUR
    1.05
     impracticable
    1.05
    할
    1.05
    ညီ
    1.04
    Act Density 0.688%

    No Known Activations