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

    Neuronpedia

    INDEX
    Explanations

    instances where the verb "haven't" or "haven" is used

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "haven" to indicate something that has not occurred or been done

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    otype
    -0.70
    dress
    -0.67
    urat
    -0.66
    upp
    -0.65
    repre
    -0.60
     fly
    -0.59
    rophe
    -0.56
    transform
    -0.55
    eller
    -0.55
    ugu
    -0.54
    POSITIVE LOGITS
    't
    1.16
    ited
    0.93
    tyard
    0.89
    geon
    0.88
     Been
    0.83
    ÃŃ
    0.82
    ãĤ§
    0.82
    dayName
    0.81
    ajor
    0.80
    iting
    0.76
    Act Density 0.025%

    No Known Activations