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

    Neuronpedia

    INDEX
    Explanations

    attends to numerical tokens from citations or references within the text

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.06
    1:0.09
    2:0.08
    3:0.07
    4:0.07
    5:0.05
    6:0.13
    7:0.40
    Negative Logits
    ThroughAttribute
    -0.57
    DockStyle
    -0.53
     المعيارى
    -0.50
    TagMode
    -0.46
     Audiodateien
    -0.44
     ostavi
    -0.43
    awtextra
    -0.41
     utafitiHapana
    -0.40
    addPreferredGap
    -0.39
    richTextPanel
    -0.39
    POSITIVE LOGITS
     ajuns
    0.29
    ("}\
    0.23
    paddingVertical
    0.23
    neros
    0.23
    )";
    0.23
    omnie
    0.23
    "/",
    0.22
    wrights
    0.22
    féle
    0.21
    რა
    0.21
    Act Density 0.170%

    No Known Activations