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

    Neuronpedia

    INDEX
    Explanations

    confirming language abilities, of course

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    いて
    0.41
    んば
    0.40
     혹
    0.40
    *((
    0.39
    的东西
    0.38
    竟
    0.37
    阅读
    0.37
    وني
    0.37
    来
    0.37
    组织的
    0.37
    POSITIVE LOGITS
     당연
    0.68
     natürlich
    0.61
    當然
    0.60
     oczywiście
    0.60
     tentunya
    0.57
     évidemment
    0.57
     tentu
    0.56
    当然
    0.55
     obviously
    0.55
     Конечно
    0.55
    Act Density 0.013%

    No Known Activations