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

    Neuronpedia

    INDEX
    Explanations

    specific instances of concepts

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     물론
    1.03
     appalling
    1.02
     Ironically
    0.95
     even
    0.93
     zelfs
    0.89
    更是
    0.89
     오히려
    0.88
     homophobic
    0.87
     Admittedly
    0.87
     übrigens
    0.87
    POSITIVE LOGITS
    某个
    1.00
    某一
    0.95
    または
    0.95
    Usually
    0.90
    किसी
    0.89
     suatu
    0.88
    通常
    0.87
     или
    0.83
     déterminé
    0.83
     одного
    0.80
    Act Density 0.645%

    No Known Activations