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

    Neuronpedia

    INDEX
    Explanations

    wrongness and incorrectness

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    선
    0.39
     changes
    0.37
     eels
    0.37
    Necessary
    0.37
    רץ
    0.35
    isnan
    0.35
     textAppearance
    0.35
    %{
    0.35
    േണ്ട
    0.35
     disturbances
    0.35
    POSITIVE LOGITS
     incorrectly
    0.72
    incorrect
    0.72
     Incorrect
    0.70
     incorrect
    0.69
    误
    0.68
     wrongly
    0.66
     неправи
    0.65
    誤
    0.65
     wrong
    0.64
     गलत
    0.64
    Act Density 0.028%

    No Known Activations