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

    Neuronpedia

    INDEX
    Explanations

    attends to list item tokens from the closing list item token

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.10
    1:0.18
    2:0.09
    3:0.06
    4:0.14
    5:0.25
    6:0.06
    7:0.07
    Negative Logits
     autorytatywna
    -0.35
     stället
    -0.31
    ์ตูน
    -0.26
     racine
    -0.25
    assertNull
    -0.25
     سكانية
    -0.24
     disambiguazione
    -0.24
    execSQL
    -0.24
     aanbod
    -0.24
     кроме
    -0.24
    POSITIVE LOGITS
    ViewFeatures
    0.38
    PostExecute
    0.35
    IBOutlet
    0.35
     Mies
    0.33
    luiting
    0.32
     Iden
    0.31
    нні
    0.31
     ]];
    0.31
     Mame
    0.30
     crucified
    0.30
    Act Density 0.004%

    No Known Activations