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

    Neuronpedia

    INDEX
    Explanations

    words related to accuracy or being correct

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "accurate."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    NetMessage
    -0.88
    hov
    -0.84
    Rus
    -0.79
    neys
    -0.79
    ATA
    -0.74
    cheon
    -0.74
    hibit
    -0.73
    LAN
    -0.73
    spl
    -0.73
    ETF
    -0.73
    POSITIVE LOGITS
     accurate
    1.27
    uracy
    1.17
     inaccurate
    1.16
     accuracy
    1.16
     inacc
    1.14
     portrayal
    1.02
     depiction
    1.00
     inaccur
    0.97
     accurately
    0.94
     truthful
    0.92
    Act Density 0.006%

    No Known Activations