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

    Neuronpedia

    INDEX
    Explanations

    specific programming code line prefixes

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
     moder
    -0.87
    llah
    -0.87
     Manson
    -0.85
    atform
    -0.85
     Ubisoft
    -0.85
     Garry
    -0.84
     Pearce
    -0.82
    quished
    -0.81
     JPM
    -0.81
     Frazier
    -0.81
    POSITIVE LOGITS
    EStreamFrame
    1.24
    default
    1.21
    dict
    1.20
    gradient
    1.19
    chance
    1.14
    SOURCE
    1.14
    tro
    1.14
    Main
    1.14
    blank
    1.14
    statement
    1.13
    Act Density 6.755%

    No Known Activations