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

    Neuronpedia

    INDEX
    Explanations

    nature

    np_max-act · gemini-2.0-flash

    The neuron is primarily triggered by French subword tokens—especially those containing accented letters.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     kennenlernen
    -0.08
    deep
    -0.07
    _ma
    -0.07
     द
    -0.06
    CEPTION
    -0.06
     desper
    -0.06
    gran
    -0.06
    Border
    -0.06
    iết
    -0.06
    >All
    -0.06
    POSITIVE LOGITS
    .group
    0.07
    WillAppear
    0.06
     NW
    0.06
    (source
    0.06
     Habitat
    0.06
    .Ge
    0.06
    .Render
    0.06
    RenderingContext
    0.06
    _tpl
    0.06
    NSS
    0.06
    Act Density 0.047%

    No Known Activations