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

    Neuronpedia

    INDEX
    Explanations

    Remote locations

    np_max-act · gemini-2.0-flash

    This neuron fires on comparative constructions that state “more X … than Y,” i.e. phrases expressing that one thing is more of some quality than another.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    namespace
    -0.07
     samen
    -0.07
    りに
    -0.07
    족
    -0.07
    _hd
    -0.07
    
    -0.06
    etě
    -0.06
    합
    -0.06
     tạo
    -0.06
    [W
    -0.06
    POSITIVE LOGITS
    .Complete
    0.06
    SECOND
    0.06
    _OPTIONS
    0.06
     VE
    0.06
     जल
    0.06
     Abb
    0.06
     CE
    0.05
    cono
    0.05
     KH
    0.05
     Courage
    0.05
    Act Density 0.004%

    No Known Activations