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

    Neuronpedia

    INDEX
    Explanations

    attends to numeric digits from tokens referencing them in brackets

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.11
    1:0.14
    2:0.13
    3:0.06
    4:0.05
    5:0.40
    6:0.03
    7:0.04
    Negative Logits
    chaun
    -0.29
     للمعارف
    -0.29
     esternos
    -0.28
    ContentAsync
    -0.28
    OBS
    -0.28
     المعيارى
    -0.28
     EconPapers
    -0.26
    ligente
    -0.26
    észetes
    -0.24
    =?";
    -0.24
    POSITIVE LOGITS
    runApp
    0.39
    AddTagHelper
    0.36
    IntoConstraints
    0.36
     CreateTagHelper
    0.35
    migrationBuilder
    0.33
    Serde
    0.33
    ReusableCell
    0.32
    RTGC
    0.32
     Rptr
    0.32
    Rptr
    0.31
    Act Density 0.017%

    No Known Activations