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

    Neuronpedia

    INDEX
    Explanations

    whatever follows it

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     вси
    0.98
     کل
    0.93
    Everything
    0.92
    すべて
    0.91
    OfWeek
    0.90
     дуже
    0.90
    つまり
    0.90
     매우
    0.89
    とても
    0.89
     همیشه
    0.88
    POSITIVE LOGITS
     venga
    1.07
     may
    1.04
     seems
    1.01
     decides
    0.97
    seems
    0.96
     might
    0.95
     happens
    0.92
    !!!"
    0.90
     ensues
    0.90
    !"
    0.90
    Act Density 0.126%

    No Known Activations