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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating incorrect or inaccurate information or understanding

    oai_token-act-pair · gpt-3.5-turbo

    instances of mischaracterization or misrepresentation

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    iaries
    -0.76
    urable
    -0.72
    tar
    -0.70
    winner
    -0.68
    cedes
    -0.67
    Lago
    -0.66
    iary
    -0.66
    hya
    -0.66
    zens
    -0.66
    contained
    -0.65
    POSITIVE LOGITS
     underest
    0.74
     mistaken
    0.74
     impression
    0.74
    é¾įå
    0.73
     mistake
    0.72
     NX
    0.71
     mistakes
    0.70
     perceptions
    0.68
    Newsletter
    0.68
     Poles
    0.68
    Act Density 0.231%

    No Known Activations