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

    Neuronpedia

    INDEX
    Explanations

    traditional methods or formats

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    of
    1.02
     relacion
    1.02
    ни
    0.97
    つ
    0.91
    ă
    0.91
     vět
    0.88
    ку
    0.87
     énorm
    0.87
    يا
    0.86
    し
    0.86
    POSITIVE LOGITS
     I
    1.39
    ir
    1.24
    us
    1.16
    ang
    1.13
    traditional
    1.10
     Been
    1.09
     Traditional
    1.02
    aditional
    1.00
    k
    0.98
    al
    0.95
    Act Density 0.017%

    No Known Activations