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

    Neuronpedia

    INDEX
    Explanations

    code

    np_max-act · gemini-2.0-flash

    The neuron activates on tokens representing file extensions (strings beginning with a dot, like “.docx”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     списка
    -0.06
    NETWORK
    -0.06
     novelist
    -0.06
     widths
    -0.06
    лись
    -0.06
    rapid
    -0.06
     lø
    -0.06
     ql
    -0.06
     MATRIX
    -0.06
     궁금
    -0.06
    POSITIVE LOGITS
     Compilation
    0.07
    ogr
    0.07
    -pointer
    0.07
     ภ
    0.07
     DRIVER
    0.06
     Discipline
    0.06
    rove
    0.06
     vàng
    0.06
    >,
    0.06
    (border
    0.06
    Act Density 0.054%

    No Known Activations