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

    Neuronpedia

    INDEX
    Explanations

    legal cases

    np_max-act · gemini-2.0-flash

    The neuron is highly sensitive to numeric tokens—especially decimal numbers and citation-style volume/page references—i.e. it detects legal citation numbers.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     chest
    -0.08
     بپ
    -0.07
     dependable
    -0.07
    finite
    -0.06
    Thread
    -0.06
    بینی
    -0.06
    vní
    -0.06
    Supplier
    -0.06
     encrypted
    -0.06
     fro
    -0.06
    POSITIVE LOGITS
     ingres
    0.07
     Award
    0.06
    ERENCE
    0.06
     Boiler
    0.06
    TAB
    0.06
    _screen
    0.06
    _message
    0.06
     wildfires
    0.06
    ILITIES
    0.06
    اجع
    0.06
    Act Density 0.001%

    No Known Activations