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

    Neuronpedia

    INDEX
    Explanations

    numerical values written in a specific format

    oai_token-act-pair · gpt-3.5-turbo

    numerical data and specific symbols in a dataset

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     sou
    -0.79
    inventoryQuantity
    -0.77
     Sou
    -0.77
     Tou
    -0.69
     Hover
    -0.69
     Ange
    -0.68
     scaven
    -0.68
     chang
    -0.68
     transf
    -0.68
    EStreamFrame
    -0.68
    POSITIVE LOGITS
    54
    1.30
    55
    1.29
    47
    1.28
    48
    1.27
    64
    1.27
    53
    1.27
    63
    1.26
    61
    1.25
    46
    1.25
    49
    1.25
    Act Density 0.127%

    No Known Activations