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

    Neuronpedia

    INDEX
    Explanations

    it takes

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Are
    0.47
    are
    0.44
    它是
    0.42
    rens
    0.40
     robust
    0.40
     effic
    0.40
    ятся
    0.39
     iteratively
    0.39
     rigorously
    0.39
     rectangles
    0.38
    POSITIVE LOGITS
     turned
    0.65
     turns
    0.65
     rained
    0.62
     sounded
    0.61
    asca
    0.61
     sorprendente
    0.59
     dawned
    0.59
     beho
    0.59
     Turned
    0.58
    turned
    0.58
    Act Density 0.058%

    No Known Activations