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

    Neuronpedia

    INDEX
    Explanations

    avoiding or exceptions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     বাসা
    0.48
     loot
    0.47
     στους
    0.44
     nesses
    0.44
     familiar
    0.43
    cosa
    0.43
     penultimate
    0.42
     mara
    0.42
     dạy
    0.41
    তারা
    0.41
    POSITIVE LOGITS
    气
    0.56
    Transparency
    0.49
    一点
    0.49
    太阳
    0.48
    透明
    0.46
    避
    0.45
    ycz
    0.45
    Compose
    0.45
    鸡蛋
    0.45
    一
    0.45
    Act Density 0.001%

    No Known Activations