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

    Neuronpedia

    INDEX
    Explanations

    mathematical fractions and equations

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    il
    0.84
    x
    0.72
    k
    0.70
    ↵
    0.58
    ש
    0.53
    el
    0.52
    ين
    0.51
     formar
    0.48
    j
    0.48
     desist
    0.47
    POSITIVE LOGITS
    {
    0.89
     a
    0.77
     be
    0.74
    0
    0.64
     হইয়৷
    0.61
    Хо
    0.59
    ۔
    0.58
     тебя
    0.55
    \
    0.55
     as
    0.55
    Act Density 0.120%

    No Known Activations