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

    Neuronpedia

    INDEX
    Explanations

    attends to tokens indicating additions or further information from tokens that specify a contrasting or complementary context

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.36
    1:0.20
    2:0.10
    3:0.05
    4:0.04
    5:0.01
    6:0.05
    7:0.15
    Negative Logits
    addContainerGap
    -0.31
    jectories
    -0.26
     Pickles
    -0.26
    ConstraintMaker
    -0.25
     Dord
    -0.25
     متعلقه
    -0.25
     aikaa
    -0.25
    øde
    -0.25
     endometrial
    -0.25
     colectiva
    -0.25
    POSITIVE LOGITS
    ⓧ
    0.43
    Hauptartikel
    0.37
    Tembelea
    0.32
     الحره
    0.32
    IsContent
    0.31
    VIER
    0.30
    CppMethod
    0.30
    QMetaType
    0.28
    AnchorTagHelper
    0.28
    fillType
    0.28
    Act Density 0.344%

    No Known Activations