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

    Neuronpedia

    INDEX
    Explanations

    attends to tokens expressing a viewpoint or a statement from earlier appearing tokens

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.08
    1:0.09
    2:0.07
    3:0.25
    4:0.18
    5:0.08
    6:0.14
    7:0.08
    Negative Logits
     AttributeSet
    -0.42
    TabIndex
    -0.40
    LookAnd
    -0.40
     Waray
    -0.40
     pinulongan
    -0.39
     BoxFit
    -0.38
    UrlResolution
    -0.37
    disposing
    -0.37
     Signalez
    -0.37
    <%@
    -0.36
    POSITIVE LOGITS
    nino
    0.36
    Tribute
    0.35
     homenaje
    0.34
    frid
    0.33
    vell
    0.32
    Lob
    0.32
     TAMBIÉN
    0.32
    culosis
    0.32
     palla
    0.32
    pelo
    0.32
    Act Density 0.091%

    No Known Activations