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

    Neuronpedia

    INDEX
    Explanations

    Code

    np_max-act · gemini-2.0-flash

    This neuron detects file‐include or import directives specifying C++ header (or source) filenames.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Compound
    -0.07
    мон
    -0.06
     자료
    -0.06
     Math
    -0.06
     yii
    -0.06
     атмос
    -0.06
     prices
    -0.06
     array
    -0.06
     blends
    -0.05
     Ep
    -0.05
    POSITIVE LOGITS
    	INT
    0.07
    .sd
    0.07
    INTERNAL
    0.07
    UpInside
    0.07
     uten
    0.07
    POR
    0.07
    idency
    0.07
    .relu
    0.07
    ��
    0.07
     tři
    0.06
    Act Density 0.015%

    No Known Activations