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

    Neuronpedia

    INDEX
    Explanations

    +

    np_max-act · gemini-2.0-flash

    The neuron is specialized to detect numeric tokens, especially floating-point numbers and scientific‐notation components (e.g. “1.23”, “E”, exponents).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    .Write
    -0.07
    .trip
    -0.06
     Sour
    -0.06
     Seek
    -0.06
    şt
    -0.06
    .Y
    -0.06
     Hüs
    -0.06
    تك
    -0.05
    ležit
    -0.05
    Inter
    -0.05
    POSITIVE LOGITS
    地區
    0.07
     PAGE
    0.07
     defence
    0.06
    ."'
    0.06
     turbine
    0.06
     &'
    0.06
    POSITORY
    0.06
    ItemType
    0.06
     acompañ
    0.06
    目标
    0.06
    Act Density 0.002%

    No Known Activations