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

    Neuronpedia

    INDEX
    Explanations

    set/create/replace/text home

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     THEY
    0.87
    They
    0.73
     굉장히
    0.70
     They
    0.70
     COULD
    0.68
    ↵↵
    0.66
    很大
    0.64
     считают
    0.64
    they
    0.63
    ?).
    0.63
    POSITIVE LOGITS
     appropriate
    1.73
     necessary
    1.59
    必要的
    1.57
     Appropriate
    1.54
    appropriate
    1.37
     Necessary
    1.35
     आवश्यक
    1.34
     notwend
    1.25
     gerekli
    1.24
     необходимые
    1.24
    Act Density 0.188%

    No Known Activations