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

    Neuronpedia

    INDEX
    Explanations

    attends to the numerical indicators from arbitrary closing tokens

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.19
    1:0.13
    2:0.11
    3:0.05
    4:0.06
    5:0.04
    6:0.06
    7:0.32
    Negative Logits
     kasarigan
    -0.52
    ViewImports
    -0.46
    )*/
    -0.43
    Datuak
    -0.42
     виправивши
    -0.41
    )";
    -0.41
    __(/*!
    -0.40
    .")]
    -0.40
     AssemblyVersion
    -0.40
    原始内容存档于
    -0.39
    POSITIVE LOGITS
    en
    0.28
    ays
    0.28
    Click
    0.26
    z
    0.25
     Click
    0.24
    .
    0.23
    ent
    0.23
     `
    0.23
     bianche
    0.23
    qu
    0.22
    Act Density 0.008%

    No Known Activations