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

    Neuronpedia

    INDEX
    Explanations

    quantiles and percentages

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    s
    1.07
    d
    0.97
    Пре
    0.91
    r
    0.87
    Ча
    0.85
    in
    0.84
    is
    0.82
    ים
    0.82
    Ма
    0.80
    र
    0.79
    POSITIVE LOGITS
    <0x80>
    1.20
    0
    1.17
    ва
    0.84
    _
    0.84
    কে
    0.79
     describir
    0.76
    ية
    0.75
    .
    0.75
    ्य
    0.73
    ある
    0.73
    Act Density 0.565%

    No Known Activations