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

    Neuronpedia

    INDEX
    Explanations

    sentences enclosed in quotation marks

    oai_token-act-pair · gpt-3.5-turbo

    quotation marks indicating spoken dialogue

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     prelim
    -0.81
     prec
    -0.73
     brid
    -0.70
     adjud
    -0.69
     leaking
    -0.69
     studio
    -0.68
     streamed
    -0.68
     streaming
    -0.68
     priesthood
    -0.67
     altar
    -0.67
    POSITIVE LOGITS
    Certainly
    1.17
    We
    1.11
    Really
    1.11
    Of
    1.07
    Obviously
    1.07
    Quite
    1.07
    Clearly
    1.05
    Without
    1.05
    That
    1.05
    Indeed
    1.04
    Act Density 0.109%

    No Known Activations