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

    Neuronpedia

    INDEX
    Explanations

    address information

    np_max-act · gemini-2.0-flash

    This neuron detects address components and related punctuation, such as street numbers and street‐type abbreviations (e.g. “Ave,” “Drive,” “Road,” “Blvd,” “Room,” “Suite”) in affiliation or contact information blocks.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    fd
    -0.07
    -sign
    -0.07
    _led
    -0.07
     inside
    -0.07
    -vector
    -0.06
    显
    -0.06
    Her
    -0.06
    .ot
    -0.06
    iman
    -0.06
     Alloy
    -0.06
    POSITIVE LOGITS
     판매
    0.07
    AccessorType
    0.07
    (KeyEvent
    0.06
    numeric
    0.06
    сутств
    0.06
    .ImageField
    0.06
    =username
    0.06
    สมบ
    0.06
    _Search
    0.06
     KeyEvent
    0.06
    Act Density 0.022%

    No Known Activations