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

    Neuronpedia

    INDEX
    Explanations

    attends to tokens indicating a relationship of increase or dependence from tokens expressing a measure or limit

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.10
    1:0.12
    2:0.11
    3:0.08
    4:0.04
    5:0.04
    6:0.07
    7:0.40
    Negative Logits
     vuitton
    -0.30
     bezeichneter
    -0.27
     Upper
    -0.24
    pely
    -0.23
     Regio
    -0.23
     beginnetje
    -0.23
     bo
    -0.23
    kover
    -0.22
    uramente
    -0.22
     Mainly
    -0.22
    POSITIVE LOGITS
     onCancelled
    0.35
    UnsafeEnabled
    0.34
    Chham
    0.34
    ticides
    0.32
     EventArgs
    0.32
     FormBuilder
    0.31
     Paglinawan
    0.31
     imprimer
    0.30
    šanās
    0.30
     CWE
    0.30
    Act Density 0.754%

    No Known Activations