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

    Neuronpedia

    INDEX
    Explanations

    attends to the token "to" from various types of tokens, including punctuation and auxiliary verbs

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.11
    1:0.12
    2:0.42
    3:0.07
    4:0.03
    5:0.02
    6:0.04
    7:0.15
    Negative Logits
     مرئيه
    -0.32
    Portail
    -0.28
    redd
    -0.27
     Vedi
    -0.27
     Thiel
    -0.27
     Lomb
    -0.26
     inflama
    -0.25
    sser
    -0.24
    re
    -0.24
     whereupon
    -0.24
    POSITIVE LOGITS
    AddTagHelper
    0.45
    SharedDtor
    0.42
    ///</
    0.39
     للاسماء
    0.39
     ProtoMessage
    0.37
    ItemBackground
    0.37
    AndEndTag
    0.36
    openConnection
    0.35
    parametrize
    0.35
    :'/
    0.35
    Act Density 0.684%

    No Known Activations