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

    Neuronpedia

    INDEX
    Explanations

    attends to sequences containing commas from tokens that appear later or are themselves tokens

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.08
    1:0.08
    2:0.57
    3:0.07
    4:0.04
    5:0.04
    6:0.02
    7:0.05
    Negative Logits
    RegistryLite
    -0.58
    цездатний
    -0.54
    interopRequire
    -0.54
     nahilalakip
    -0.52
     pageContext
    -0.50
     OFDb
    -0.49
     Seeder
    -0.49
     الاطلاع
    -0.47
     transfieras
    -0.47
    oneofs
    -0.47
    POSITIVE LOGITS
     ambos
    0.37
     keduanya
    0.32
    Closeable
    0.32
    ientôt
    0.31
     दोनों
    0.31
     beide
    0.31
     dupla
    0.30
    textTheme
    0.30
    leder
    0.29
     Beide
    0.29
    Act Density 0.782%

    No Known Activations