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

    Neuronpedia

    INDEX
    Explanations

    attends to punctuation tokens from tokens indicating the context of code and data handling

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.16
    1:0.14
    2:0.12
    3:0.09
    4:0.06
    5:0.03
    6:0.06
    7:0.32
    Negative Logits
    WithIOException
    -0.51
    SharedDtor
    -0.48
    StoryboardSegue
    -0.44
    EDEFAULT
    -0.40
    ScopeManager
    -0.39
     ModelExpression
    -0.39
    AndEndTag
    -0.38
    Rüyada
    -0.37
    usermodel
    -0.37
     שוליים
    -0.37
    POSITIVE LOGITS
    Насе
    0.23
    Param
    0.21
    They
    0.21
    Bang
    0.21
    av
    0.21
    sted
    0.20
     dite
    0.20
    BN
    0.20
     kiedy
    0.20
    ab
    0.20
    Act Density 0.209%

    No Known Activations