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

    Neuronpedia

    INDEX
    Explanations

    exactness or specificity in statements or requests

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "exactly."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     Tome
    -0.75
    rift
    -0.73
    roe
    -0.70
    kj
    -0.70
    ker
    -0.68
    qs
    -0.68
    oly
    -0.66
    bably
    -0.66
    jay
    -0.65
    rouse
    -0.62
    POSITIVE LOGITS
    ifications
    0.73
    usha
    0.69
     HOW
    0.68
    abouts
    0.67
     how
    0.66
     amount
    0.65
    itude
    0.64
     whereabouts
    0.62
     precise
    0.62
     exactly
    0.62
    Act Density 0.018%

    No Known Activations