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

    Neuronpedia

    INDEX
    Explanations

    approximation

    np_max-act · gemini-2.0-flash

    The neuron activates on numeric expressions (especially proportions, fractions, and decimal figures).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    tabpanel
    -0.07
    ظمة
    -0.07
    lys
    -0.06
     hâl
    -0.06
    Detach
    -0.06
     Kenny
    -0.06
     Tablets
    -0.06
     cellar
    -0.06
    Composer
    -0.06
     fasting
    -0.06
    POSITIVE LOGITS
     此
    0.07
     mileage
    0.07
    ностью
    0.06
    Altern
    0.06
    .uri
    0.06
    ,)↵
    0.06
    /conf
    0.06
    /admin
    0.06
    ordinates
    0.06
     concealed
    0.06
    Act Density 0.013%

    No Known Activations