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

    Neuronpedia

    INDEX
    Explanations

    words related to weaknesses and flaws, especially in a critical context

    oai_token-act-pair · gpt-3.5-turbo

    terms related to flaws, weaknesses, and deficiencies

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    rollers
    -0.90
    iversary
    -0.82
    ctors
    -0.73
    winner
    -0.70
    ramid
    -0.70
    amaz
    -0.69
    cise
    -0.68
    da
    -0.68
    urai
    -0.66
    rooms
    -0.65
    POSITIVE LOGITS
     inherent
    1.12
     flaws
    1.05
     plag
    1.03
     weaknesses
    0.97
     vulnerabilities
    0.88
     deficiencies
    0.87
     flaw
    0.86
    ibilities
    0.86
    ulner
    0.78
     underlying
    0.77
    Act Density 0.129%

    No Known Activations