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

    Neuronpedia

    INDEX
    Explanations

    attends to numeric tokens marked with specific values from tokens that serve as contextual anchors

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.06
    1:0.11
    2:0.07
    3:0.08
    4:0.43
    5:0.08
    6:0.07
    7:0.07
    Negative Logits
     @"/
    -0.27
    deque
    -0.25
     Ahl
    -0.25
    anks
    -0.24
     gỗ
    -0.23
    addLast
    -0.22
    raster
    -0.22
    contentLoaded
    -0.22
    bitan
    -0.22
    ://
    -0.21
    POSITIVE LOGITS
    RenderAtEndOf
    0.38
    WriteBarrier
    0.35
     Exacts
    0.34
    ActionCreators
    0.33
    exitRule
    0.32
    setVerticalGroup
    0.31
    ScopeManager
    0.30
    oa̍t
    0.29
     ());
    0.29
    InitVars
    0.29
    Act Density 0.875%

    No Known Activations