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

    Neuronpedia

    INDEX
    Explanations

    avoid bias then develop

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     ойноо
    0.30
     акчага
    0.30
     ойношот
    0.29
    那我們
    0.28
    Ты
    0.28
     аўтаматы
    0.27
     vattati
    0.27
     ойной
    0.27
    Benzyl
    0.27
     vuccanti
    0.27
    POSITIVE LOGITS
    可以
    0.37
    在
    0.35
    有
    0.33
    更
    0.32
    选择
    0.32
    不
    0.32
    能
    0.32
    直接
    0.32
    对
    0.31
    做
    0.31
    Act Density 0.070%

    No Known Activations