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

    Neuronpedia

    INDEX
    Explanations

    combinations of words related to tools, construction, and procedures

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
     Caption
    -0.73
    ffiti
    -0.59
    erer
    -0.58
    resy
    -0.58
    Ł
    -0.57
    ÅĤ
    -0.57
    zbollah
    -0.57
    ario
    -0.56
    hran
    -0.56
    gun
    -0.56
    POSITIVE LOGITS
     sorts
    0.78
     disparate
    0.73
     styles
    0.72
     genres
    0.71
     sexes
    0.69
     halves
    0.66
     disciplines
    0.65
     goodies
    0.62
     flavors
    0.62
     positives
    0.60
    Act Density 13.866%

    No Known Activations