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

    Neuronpedia

    INDEX
    Explanations

    attends to closing interrogative sentences marked by question marks from later tokens in the sentences

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.06
    1:0.08
    2:0.07
    3:0.11
    4:0.11
    5:0.05
    6:0.38
    7:0.09
    Negative Logits
    Pautan
    -0.41
     Italijani
    -0.41
    лтамалар
    -0.41
     שוליים
    -0.40
     "..\..\..\
    -0.40
     متعلقه
    -0.39
     Waray
    -0.38
    liken
    -0.38
    Diweddarwch
    -0.38
     հղումներ
    -0.37
    POSITIVE LOGITS
     Vordergrund
    0.36
    öny
    0.34
    Tikang
    0.34
     excru
    0.32
     commenting
    0.32
    matsu
    0.31
    سب
    0.31
    dorfer
    0.30
    NotExist
    0.30
    marily
    0.30
    Act Density 0.008%

    No Known Activations