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

    Neuronpedia

    INDEX
    Explanations

    attends to numeric values from associated percentages or measurements

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.14
    1:0.18
    2:0.10
    3:0.08
    4:0.07
    5:0.03
    6:0.07
    7:0.29
    Negative Logits
    )++;
    -0.34
    ."\
    -0.30
     --;
    -0.30
    (":");
    -0.29
     joyeux
    -0.29
    romptu
    -0.29
    kään
    -0.29
     چیه
    -0.29
     neceffary
    -0.29
     Zeitung
    -0.29
    POSITIVE LOGITS
    RenderAtEndOf
    0.44
    ReusableCell
    0.44
    MemoryWarning
    0.43
    abestanden
    0.43
    DataAnnotations
    0.42
    endregion
    0.42
     оригіналу
    0.41
    OGND
    0.39
     Meksiku
    0.39
    InjectAttribute
    0.38
    Act Density 0.514%

    No Known Activations