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

    Neuronpedia

    INDEX
    Explanations

    action descriptions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     underscore
    0.46
    ’)
    0.45
    ’
    0.45
     consente
    0.45
     print
    0.45
     are
    0.44
    ”)
    0.44
     paddock
    0.43
     prints
    0.42
     discrepancy
    0.42
    POSITIVE LOGITS
    April
    0.47
    当时
    0.47
    May
    0.47
    と思いました
    0.45
    𝐁
    0.45
    閱讀
    0.45
     ايه
    0.44
    ?(
    0.44
    Hope
    0.44
    Это
    0.43
    Act Density 0.006%

    No Known Activations