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

    Neuronpedia

    INDEX
    Explanations

    Type

    np_max-act · gemini-2.0-flash

    This neuron detects German word stems and suffix pieces that form abstract nouns—especially segments like “Veränder-,” “physischen Bedingungen,” “Umgebung,” and “Realität” (i.e. “-ung” and “-ität” endings).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    ์บ
    -0.07
    ANN
    -0.07
     TMP
    -0.06
    LER
    -0.06
     Mac
    -0.06
     FA
    -0.06
     beaches
    -0.06
    operator
    -0.06
    sut
    -0.06
    Fetching
    -0.06
    POSITIVE LOGITS
     Rhode
    0.08
    .setHorizontalGroup
    0.07
    _screen
    0.07
    -photo
    0.07
     says
    0.06
     composition
    0.06
     ağır
    0.06
     "\↵
    0.06
    .graph
    0.06
     nx
    0.06
    Act Density 0.002%

    No Known Activations