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

    Neuronpedia

    INDEX
    Explanations

    proper nouns related to heroes or knights

    oai_token-act-pair · gpt-3.5-turbo

    mentions of characters or groups commonly associated with heroic or fantastical themes

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    izations
    -0.71
    icity
    -0.71
    lying
    -0.70
    iment
    -0.69
    ffer
    -0.68
    lier
    -0.67
    abases
    -0.66
    rique
    -0.66
    leased
    -0.65
    lied
    -0.65
    POSITIVE LOGITS
    mith
    1.20
    aurus
    1.19
    hip
    1.15
    layer
    1.08
    hift
    1.06
    cale
    1.03
    '
    0.97
    peed
    0.96
     Reborn
    0.92
    linger
    0.92
    Act Density 0.129%

    No Known Activations