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

    Neuronpedia

    INDEX
    Explanations

    ,

    np_max-act · gemini-2.0-flash

    The neuron activates on the Chinese comma-plus-“you” token (“，你”), i.e. it detects the punctuation-boundary where the speaker directly addresses “you” in Mandarin.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    ैय
    -0.07
     septembre
    -0.07
    take
    -0.07
    小姐
    -0.06
     attravers
    -0.06
    Achie
    -0.06
     dementia
    -0.06
     dilation
    -0.06
     Cele
    -0.06
    �
    -0.06
    POSITIVE LOGITS
    iqueta
    0.07
    _FT
    0.06
    恒
    0.06
    较
    0.06
    pective
    0.06
     normally
    0.06
     Каб
    0.06
     vertices
    0.06
     للإ
    0.06
     Глав
    0.06
    Act Density 0.000%

    No Known Activations