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

    Neuronpedia

    INDEX
    Explanations

    normalizes and tailors

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     the
    0.50
     using
    0.49
     an
    0.44
     adapting
    0.44
     during
    0.43
     their
    0.43
     administrative
    0.42
     with
    0.41
     are
    0.41
     adapted
    0.41
    POSITIVE LOGITS
    在这里
    0.47
    也会
    0.45
    枼
    0.44
    赎
    0.42
    劻
    0.42
    不一样
    0.42
    诈
    0.41
    Ở
    0.41
     tutaj
    0.41
     trochę
    0.41
    Act Density 0.054%

    No Known Activations