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

    Neuronpedia

    INDEX
    Explanations

    And then

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    但在
    0.50
    但
    0.46
     ancak
    0.45
     Однако
    0.43
     azonban
    0.43
    但是在
    0.42
     ;,
    0.41
     pero
    0.41
    কিন্ত
    0.40
    endering
    0.40
    POSITIVE LOGITS
     importantly
    0.64
     speaking
    0.61
     lastly
    0.61
    rée
    0.60
     yes
    0.59
     кстати
    0.59
    ंगाबाद
    0.58
     crucially
    0.57
    ean
    0.56
    termilk
    0.56
    Act Density 0.021%

    No Known Activations