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

    Neuronpedia

    INDEX
    Explanations

    consistently even

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    少了
    0.73
     azar
    0.64
    ষম
    0.63
     volonté
    0.62
     délai
    0.61
     five
    0.61
    )/
    0.60
     homog
    0.60
    effic
    0.60
     two
    0.59
    POSITIVE LOGITS
    甚至
    1.60
     bahkan
    1.58
     Bahkan
    1.48
     zelfs
    1.43
     thậm
    1.41
     even
    1.39
     nawet
    1.32
    甚至是
    1.31
     sogar
    1.30
     addirittura
    1.28
    Act Density 0.654%

    No Known Activations