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

    Neuronpedia

    INDEX
    Explanations

    referring to previous or earlier things

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     traditional
    0.43
    伝統
    0.41
    懐
    0.41
     old
    0.41
     simply
    0.39
     tradicional
    0.38
     actual
    0.38
     semplicemente
    0.38
     일단
    0.37
     Traditional
    0.37
    POSITIVE LOGITS
     earlier
    2.69
     Earlier
    2.41
    Earlier
    2.36
     sebelumnya
    2.33
     tidigare
    2.31
     tidligere
    2.27
     ранее
    2.23
     wcześniej
    2.20
     zuvor
    2.17
     wcześ
    2.13
    Act Density 0.040%

    No Known Activations