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

    Neuronpedia

    INDEX
    Explanations

    attends to key nouns related to specific subjects from corresponding contextually relevant tokens

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Head Attr Weights
    0:0.08
    1:0.08
    2:0.06
    3:0.08
    4:0.39
    5:0.10
    6:0.09
    7:0.07
    Negative Logits
    Rptr
    -0.29
    bardier
    -0.28
     frapp
    -0.26
     Hars
    -0.25
    fdf
    -0.25
    Données
    -0.25
    FV
    -0.25
    पया
    -0.25
     espuma
    -0.25
     Trades
    -0.24
    POSITIVE LOGITS
    ExecuteReader
    0.42
    ThroughAttribute
    0.41
     NDEBUG
    0.41
    styleable
    0.40
    WebControls
    0.39
    BeginContext
    0.36
     Normdatei
    0.35
    httphttps
    0.35
    ftagPool
    0.35
     OnTriggerEnter
    0.35
    Act Density 0.296%

    No Known Activations