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

    Neuronpedia

    INDEX
    Explanations

    attends to numerical changes indicated by "to" from subsequent numerical tokens

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.12
    1:0.15
    2:0.12
    3:0.06
    4:0.06
    5:0.03
    6:0.05
    7:0.37
    Negative Logits
     насељу
    -0.37
    ʒ
    -0.35
     Barbuda
    -0.34
     Gleaner
    -0.33
     EnglishChoose
    -0.30
    secta
    -0.30
     kterou
    -0.29
    
    -0.29
    CreateModel
    -0.29
    __':
    -0.29
    POSITIVE LOGITS
    ReusableCell
    0.41
    IntoConstraints
    0.29
    存于互联网档案馆
    0.29
    ใจ
    0.28
    makeConstraints
    0.28
    addContainerGap
    0.27
    unary
    0.26
    beginPath
    0.26
    PostMapping
    0.25
    zaj
    0.24
    Act Density 0.089%

    No Known Activations