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

    Neuronpedia

    INDEX
    Explanations

    attends to a capitalized token from a later lowercased token or a preceding capitalized token

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.07
    1:0.10
    2:0.09
    3:0.10
    4:0.09
    5:0.05
    6:0.27
    7:0.20
    Negative Logits
     amante
    -0.23
    padek
    -0.23
     medes
    -0.23
     able
    -0.23
     ää
    -0.22
    gaande
    -0.22
     
    -0.21
    ści
    -0.21
     propOrder
    -0.21
     Roskov
    -0.20
    POSITIVE LOGITS
    DockStyle
    0.54
     '\\;'
    0.43
    ganu
    0.43
    UrlResolution
    0.38
    SequentialGroup
    0.38
     ſind
    0.38
    ArgsConstructor
    0.38
    RenderAtEndOf
    0.38
    IBarButtonItem
    0.37
     iſt
    0.37
    Act Density 0.057%

    No Known Activations