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

    Neuronpedia

    INDEX
    Explanations

    News articles/blog posts

    np_max-act · gemini-2.0-flash

    The neuron fires on named entities—especially organization or brand names (e.g. “Yahoo,” “Facebook Connect,” “Nigeria Premier League”)—flagging tokens that are part of proper nouns.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     forestry
    -0.06
     respecto
    -0.06
     plug
    -0.06
    oph
    -0.06
     cultures
    -0.06
    Tuple
    -0.06
     turmoil
    -0.06
    (title
    -0.06
    ahu
    -0.06
     Bentley
    -0.06
    POSITIVE LOGITS
    GetY
    0.07
    0.07
    ViewChild
    0.07
    vider
    0.07
     Predator
    0.06
     chụp
    0.06
    _DIS
    0.06
     cancelButton
    0.06
    *******/↵↵
    0.06
    unte
    0.06
    Act Density 0.112%

    No Known Activations