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

    Neuronpedia

    INDEX
    Explanations

    words that seem to have encoding errors, particularly related to the characters 'Â' and 'ĵ'

    oai_token-act-pair · gpt-3.5-turbo

    the symbol "Â" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    enegger
    -0.94
     Nanto
    -0.80
    manship
    -0.70
    anwhile
    -0.70
    ãģ®éŃĶ
    -0.69
     comprehens
    -0.67
     Kamp
    -0.67
    iants
    -0.66
     Mamm
    -0.66
     VK
    -0.66
    POSITIVE LOGITS
    ´
    1.61
    ³
    1.54
    ¿
    1.51
    ¨
    1.50
    ¦
    1.49
    ¹
    1.47
    ¬
    1.41
    ¸
    1.40
    ¢
    1.39
    ¼
    1.39
    Act Density 0.014%

    No Known Activations