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

    Neuronpedia

    INDEX
    Explanations

    capital letters with unusual accents and punctuation marks

    oai_token-act-pair · gpt-3.5-turbo

    numeric values and significant numerical statistics in discussions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    boro
    -0.73
    utics
    -0.70
     undet
    -0.69
    utic
    -0.68
     lifes
    -0.65
     hust
    -0.63
     wiser
    -0.62
    utical
    -0.61
    utsche
    -0.60
     attent
    -0.60
    POSITIVE LOGITS
    However
    0.95
    SPONSORED
    0.95
    Meanwhile
    0.95
    Advertisement
    0.94
    Newsletter
    0.92
    Article
    0.90
    Nevertheless
    0.89
    Despite
    0.88
    Related
    0.87
    Unlike
    0.87
    Act Density 1.714%

    No Known Activations