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

    Neuronpedia

    INDEX
    Explanations

    attends to offseason-related tokens from training-related tokens

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.12
    1:0.15
    2:0.11
    3:0.12
    4:0.13
    5:0.04
    6:0.13
    7:0.16
    Negative Logits
    WebpackPlugin
    -0.34
     vectorielle
    -0.33
     automatiques
    -0.33
     vectorielles
    -0.32
    harusnya
    -0.30
    égories
    -0.30
    magitan
    -0.30
    SourceChecksum
    -0.29
    UAWEI
    -0.29
     käyttö
    -0.29
    POSITIVE LOGITS
    be
    0.25
    sig
    0.24
     focusing
    0.24
    iton
    0.24
     sig
    0.24
    l
    0.24
     reac
    0.23
     Be
    0.23
     Jou
    0.23
    ANTLR
    0.23
    Act Density 0.070%

    No Known Activations