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

    Neuronpedia

    INDEX
    Explanations

    attends to the temporal term "when" from emotionally reflective or descriptive terms

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.13
    1:0.14
    2:0.09
    3:0.12
    4:0.12
    5:0.02
    6:0.16
    7:0.18
    Negative Logits
    principalColumn
    -0.27
    <bos>
    -0.26
     numerus
    -0.25
     συγ
    -0.24
    ंगा
    -0.24
    뛰
    -0.23
    urent
    -0.23
     kanan
    -0.23
     cuello
    -0.23
    acheter
    -0.23
    POSITIVE LOGITS
    Portale
    0.35
    FunctionFlags
    0.33
    GeoNames
    0.33
     Allociné
    0.32
    RefreshLayout
    0.31
    IntoConstraints
    0.30
    expandindo
    0.30
    PhysRevLett
    0.29
     betweenstory
    0.29
     Excerpt
    0.28
    Act Density 0.182%

    No Known Activations