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

    Neuronpedia

    INDEX
    Explanations

    strings that end with a colon

    oai_token-act-pair · gpt-3.5-turbo

    colons followed by commands or instructions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ĸļ
    -0.85
    undai
    -0.79
    izons
    -0.78
     tremend
    -0.74
    odox
    -0.71
    ikuman
    -0.70
    dinand
    -0.69
     subur
    -0.68
    oud
    -0.68
    etooth
    -0.67
    POSITIVE LOGITS
     namely
    0.93
    ³³³³³³³³³³³³³³³³
    0.92
     Logged
    0.90
     Provided
    0.87
     Bye
    0.85
    YES
    0.85
     http
    0.81
     Firstly
    0.79
     https
    0.77
     ie
    0.77
    Act Density 0.175%

    No Known Activations