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

    Neuronpedia

    INDEX
    Explanations

    timestamps

    np_max-act · gemini-2.0-flash

    The neuron fires primarily on numeric tokens representing decimal values (numbers containing a period and fractional part).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    .div
    -0.07
     limited
    -0.07
    008
    -0.06
    serialization
    -0.06
    .Env
    -0.06
    Cnt
    -0.06
    743
    -0.06
     slur
    -0.06
    eated
    -0.06
    364
    -0.06
    POSITIVE LOGITS
     jeux
    0.08
    gger
    0.06
     HttpHeaders
    0.06
     đột
    0.06
     Invest
    0.06
     ventas
    0.06
    ฤษ
    0.06
     Tân
    0.06
     Haiti
    0.06
     стоит
    0.06
    Act Density 0.006%

    No Known Activations