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

    Neuronpedia

    INDEX
    Explanations

    contractions

    oai_token-act-pair · gpt-3.5-turbo

    instances of the phrase "can't" or similar contractions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    çīĪ
    -0.86
     RAD
    -0.81
     Gleaming
    -0.67
    enegger
    -0.65
    代
    -0.64
    ãĤ¼
    -0.64
    æĪ¦
    -0.63
     airs
    -0.63
    76561
    -0.62
    stocks
    -0.62
    POSITIVE LOGITS
    Ķ
    1.06
    £
    1.04
    ĵ
    1.00
    ¨
    0.99
    «
    0.98
    ¼
    0.97
    lege
    0.94
    ı
    0.94
    ij
    0.94
    º
    0.93
    Act Density 0.060%

    No Known Activations