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

    Neuronpedia

    INDEX
    Explanations

    So, From

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     interestingly
    0.52
     unsurprisingly
    0.44
     consequently
    0.44
    Interestingly
    0.43
     Interestingly
    0.42
     sekaligus
    0.42
     redefine
    0.41
     أثناء
    0.41
     inten
    0.40
     numeracy
    0.40
    POSITIVE LOGITS
     Hark
    0.45
    So
    0.44
     Пусть
    0.44
     Từ
    0.43
    ug
    0.42
    From
    0.42
     Praise
    0.42
     speedily
    0.42
     Rejo
    0.42
    Though
    0.41
    Act Density 0.043%

    No Known Activations