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

    Neuronpedia

    INDEX
    Explanations

    issued

    np_max-act · gemini-2.0-flash

    The neuron activates on floating‐point numeric tokens (decimal numbers with a point and digits).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    (description
    -0.07
     olursa
    -0.06
    ponsored
    -0.06
    getTable
    -0.06
     пи
    -0.06
    Bubble
    -0.06
    (mac
    -0.06
    .live
    -0.06
     Plymouth
    -0.06
    (ad
    -0.06
    POSITIVE LOGITS
     لازم
    0.07
     مص
    0.07
     своє
    0.06
     qx
    0.06
    idf
    0.06
     Doug
    0.06
    IRQ
    0.06
     hue
    0.06
    aria
    0.06
    روع
    0.06
    Act Density 0.002%

    No Known Activations