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

    Neuronpedia

    INDEX
    Explanations

    attends to conditional tokens indicating doubt or inquiry from subsequent tokens asserting resolution or existence

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.12
    1:0.11
    2:0.11
    3:0.04
    4:0.02
    5:0.01
    6:0.09
    7:0.45
    Negative Logits
    jetunion
    -0.26
    tanleria
    -0.26
     propOrder
    -0.24
    Trials
    -0.24
    psy
    -0.24
    PreferredItem
    -0.23
    xel
    -0.23
    twimg
    -0.23
    visibility
    -0.22
     ganglion
    -0.22
    POSITIVE LOGITS
    IDATE
    0.38
     وتسجيلات
    0.38
     or
    0.36
    ?");
    0.36
     quelconque
    0.36
    TestingModule
    0.36
     eller
    0.36
    ?";
    0.36
    ?')
    0.35
    それとも
    0.35
    Act Density 0.447%

    No Known Activations