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

    Neuronpedia

    INDEX
    Explanations

    attends to programming tokens from unrelated tokens in some sequences

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.07
    1:0.10
    2:0.07
    3:0.46
    4:0.09
    5:0.03
    6:0.07
    7:0.08
    Negative Logits
     placés
    -0.41
     automatiques
    -0.40
     intérieur
    -0.39
     equipado
    -0.38
     esternos
    -0.38
     aikaa
    -0.37
     ansatte
    -0.37
    AppMethodBeat
    -0.37
     Fahrt
    -0.36
     råd
    -0.36
    POSITIVE LOGITS
    MemoryWarning
    0.46
    verwijspagina
    0.43
    IsContent
    0.40
     Biele
    0.37
     initComponents
    0.36
     Rodeo
    0.35
    '])->
    0.35
    Regressor
    0.35
    ']).
    0.35
    *********/
    0.34
    Act Density 0.417%

    No Known Activations