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

    Neuronpedia

    INDEX
    Explanations

    relatively manageable states

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Invalid
    0.86
     incomplete
    0.86
    incomplete
    0.84
     незакон
    0.82
     Invalid
    0.81
     Unable
    0.80
     неправи
    0.80
     leider
    0.78
     Incorrect
    0.78
     illegally
    0.77
    POSITIVE LOGITS
     manageable
    2.16
     managable
    2.08
     relatively
    1.86
     tolerable
    1.81
     harmless
    1.77
     benign
    1.68
    relatively
    1.68
    比較的
    1.62
     Relatively
    1.62
     milder
    1.61
    Act Density 0.556%

    No Known Activations