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

    Neuronpedia

    INDEX
    Explanations

    attends to numeric references marked from elements that are related to them

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.10
    1:0.15
    2:0.11
    3:0.11
    4:0.14
    5:0.14
    6:0.08
    7:0.13
    Negative Logits
     intéress
    -0.28
    according
    -0.28
     according
    -0.28
    According
    -0.27
    is
    -0.26
    ly
    -0.26
    j
    -0.26
     Wirk
    -0.25
     TRAV
    -0.25
    ậc
    -0.25
    POSITIVE LOGITS
     initComponents
    0.44
     consultato
    0.41
    SequentialGroup
    0.38
    ConstraintMaker
    0.36
    StoryboardSegue
    0.33
    httphttps
    0.30
    cifix
    0.30
     AssemblyTitle
    0.30
    bardier
    0.29
    erequisite
    0.29
    Act Density 0.529%

    No Known Activations