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

    Neuronpedia

    INDEX
    Explanations

    odd or even numbers

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     giri
    0.69
     مخ
    0.64
    时长
    0.64
     бюро
    0.60
    ontaneous
    0.60
     নির্ভরযোগ্য
    0.59
    長時間
    0.59
    latent
    0.59
     tolerances
    0.58
    鑫
    0.58
    POSITIVE LOGITS
     odd
    3.13
    Odd
    2.84
     Odd
    2.79
    odd
    2.58
     Even
    2.34
    Even
    2.31
     even
    2.31
    even
    2.14
    奇
    2.13
    偶
    2.11
    Act Density 0.445%

    No Known Activations