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

    Neuronpedia

    INDEX
    Explanations

    humor

    np_max-act · gemini-2.0-flash

    The neuron flags words and phrases that signal humor or parody (e.g. “parodies,” “parody,” “hilarious,” “tongue-in-cheek,” “Comic,” etc.).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Somali
    -0.07
    [val
    -0.07
    [g
    -0.07
     Dog
    -0.07
    Dog
    -0.06
     Norte
    -0.06
    (IConfiguration
    -0.06
    lahoma
    -0.06
     manga
    -0.06
     Howe
    -0.06
    POSITIVE LOGITS
     hoops
    0.06
    高中
    0.06
     laure
    0.06
    .'.$
    0.06
    _.
    0.06
    (mac
    0.06
    /menu
    0.06
    _ADV
    0.06
    -j
    0.06
     mort
    0.06
    Act Density 0.088%

    No Known Activations