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

    Neuronpedia

    INDEX
    Explanations

    attends to method calls involving "traverse" from tokens indicating conditions or fields being defined

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.06
    1:0.08
    2:0.07
    3:0.11
    4:0.07
    5:0.03
    6:0.38
    7:0.15
    Negative Logits
    ecma
    -0.40
    "?>
    -0.35
     @{
    -0.33
    ostante
    -0.32
    췄
    -0.31
    rijfs
    -0.30
    BIDDEN
    -0.30
    /*
    -0.30
    geries
    -0.30
    وأضاف
    -0.29
    POSITIVE LOGITS
     متعلقه
    0.44
    
    0.33
    ReusableCell
    0.29
     klock
    0.27
    Humor
    0.25
    StoreMessageInfo
    0.25
    مراجع
    0.25
    OptionsMenu
    0.24
    antara
    0.24
     heiligen
    0.24
    Act Density 0.005%

    No Known Activations