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

    Neuronpedia

    INDEX
    Explanations

    attends to empty patterns from unspecified tokens

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.14
    1:0.29
    2:0.09
    3:0.09
    4:0.12
    5:0.11
    6:0.05
    7:0.08
    Negative Logits
     disambiguazione
    -0.45
     photolibrary
    -0.41
     esternos
    -0.40
     autorytatywna
    -0.39
     themſelves
    -0.39
    )";
    -0.39
    ſelf
    -0.38
     myſelf
    -0.38
    AsUp
    -0.38
     ModelExpression
    -0.38
    POSITIVE LOGITS
    </h2>
    0.24
    حه
    0.23
     not
    0.23
    neq
    0.22
    $$
    0.22
    ↵
    0.22
     obra
    0.22
    ांकि
    0.22
    </u>
    0.22
    (
    0.22
    Act Density 1.222%

    No Known Activations