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

    Neuronpedia

    INDEX
    Explanations

    conversations and interactions between characters in a text

    oai_token-act-pair · gpt-3.5-turbo

    expressions of humor and light-hearted dialogue

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     NCT
    -0.71
     batters
    -0.62
    é¾įå
    -0.62
     Semin
    -0.60
    hower
    -0.60
    Stat
    -0.60
     predominantly
    -0.59
    ordes
    -0.59
    uably
    -0.59
    Mand
    -0.59
    POSITIVE LOGITS
    -"
    1.45
    âĢ¦"
    1.33
    ..."
    1.27
    —"
    1.15
     âĢ¦"
    1.15
    !?"
    1.07
    âĢ¦."
    1.06
    ?"
    1.06
    !"
    0.99
    ?!"
    0.98
    Act Density 0.424%

    No Known Activations