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

    Neuronpedia

    INDEX
    Explanations

    attends to numerical tokens from textual tokens

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.11
    1:0.17
    2:0.12
    3:0.08
    4:0.11
    5:0.12
    6:0.13
    7:0.13
    Negative Logits
    amaño
    -0.50
    SequentialGroup
    -0.49
     مرئيه
    -0.49
    loroethene
    -0.45
    ρισσότε
    -0.45
    ContentAsync
    -0.44
    harusnya
    -0.44
     GenerationType
    -0.44
    cientos
    -0.43
     ExecuteAsync
    -0.41
    POSITIVE LOGITS
    SharedDtor
    0.37
    ToServer
    0.35
     Beyer
    0.35
    camore
    0.35
     aging
    0.35
     Aging
    0.35
    cig
    0.35
     Blair
    0.35
    DebuggerNonUser
    0.35
     mold
    0.34
    Act Density 0.000%

    No Known Activations