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

    Neuronpedia

    INDEX
    Explanations

    attends to opening parentheses from closing parentheses in code-related contexts

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.15
    1:0.12
    2:0.29
    3:0.08
    4:0.08
    5:0.10
    6:0.06
    7:0.09
    Negative Logits
     utafitiHapana
    -0.43
    tagHelperRunner
    -0.37
    EndContext
    -0.35
    Diweddarwch
    -0.35
    DispatchToProps
    -0.33
    AndEndTag
    -0.31
    ')):
    -0.31
    InvalidProtocol
    -0.31
     ComVisible
    -0.30
     typelib
    -0.30
    POSITIVE LOGITS
    Facades
    0.29
    ArgumentParser
    0.27
    AIRE
    0.24
     intermedia
    0.24
    Initial
    0.23
     landen
    0.23
     soucis
    0.23
    atimes
    0.23
    initi
    0.23
    duga
    0.23
    Act Density 0.053%

    No Known Activations