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

    Neuronpedia

    INDEX
    Explanations

    deep breath before action

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     begitu
    1.02
     uguale
    0.97
    いる
    0.95
    としては
    0.95
    場合は
    0.93
    ときは
    0.93
     sebesar
    0.92
     gördüğünüz
    0.91
     quando
    0.90
     ganzen
    0.90
    POSITIVE LOGITS
     卡
    1.12
     hsize
    1.07
     拉
    1.00
     餐
    0.99
     做
    0.99
     控制
    0.96
     测试
    0.95
     prune
    0.93
    лая
    0.92
     snout
    0.92
    Act Density 0.001%

    No Known Activations