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

    Neuronpedia

    INDEX
    Explanations

    phrases relating to requests or instructions

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
     divor
    -0.58
    iolet
    -0.57
    apor
    -0.56
     satur
    -0.56
     segreg
    -0.55
    flying
    -0.54
     blat
    -0.54
    redo
    -0.54
     quadru
    -0.53
     appropri
    -0.53
    POSITIVE LOGITS
     of
    0.86
     thereof
    0.74
    OF
    0.74
    aeper
    0.72
     Of
    0.72
     OF
    0.71
    oft
    0.68
     whims
    0.67
    forts
    0.65
    liest
    0.65
    Act Density 0.207%

    No Known Activations