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

    Neuronpedia

    INDEX
    Explanations

    phrases or sentences containing the word "wrong"

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating errors or failures

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    gil
    -0.77
    ecd
    -0.70
    prise
    -0.70
    resent
    -0.69
    ocry
    -0.69
    weekly
    -0.68
    Portland
    -0.66
    cit
    -0.66
    otine
    -0.66
    urst
    -0.66
    POSITIVE LOGITS
     havoc
    0.75
     unexpectedly
    0.67
     miser
    0.65
     horribly
    0.64
     onstage
    0.64
     Oops
    0.63
    aneously
    0.62
     during
    0.62
     catast
    0.61
    eria
    0.61
    Act Density 0.039%

    No Known Activations