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

    Neuronpedia

    INDEX
    Explanations

    maintain

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
    =args
    -0.06
    -*
    -0.06
     rumors
    -0.06
    肖
    -0.06
     Params
    -0.06
     sigma
    -0.06
     roof
    -0.06
    ystery
    -0.06
    Pretty
    -0.06
     provinces
    -0.06
    POSITIVE LOGITS
     maintaining
    0.09
     maintained
    0.07
     upkeep
    0.07
     dismantle
    0.07
     maintain
    0.07
     midi
    0.07
    yleft
    0.07
     unten
    0.07
    ние
    0.07
    样
    0.06
    Act Density 0.027%

    No Known Activations