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

    Neuronpedia

    INDEX
    Explanations

    depends on conditions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     waardoor
    0.46
     sehingga
    0.44
     wodurch
    0.43
     cioè
    0.41
     perangkat
    0.40
    を使って
    0.40
     Sehingga
    0.40
     Dadurch
    0.39
     derrot
    0.39
     zodat
    0.39
    POSITIVE LOGITS
     depends
    0.50
    有两种
    0.47
    に関しては
    0.45
     YES
    0.45
    については
    0.42
     varies
    0.42
    ：
    0.42
     Interestingly
    0.41
    YES
    0.41
    Depends
    0.40
    Act Density 0.256%

    No Known Activations