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

    Neuronpedia

    INDEX
    Explanations

    year numbers

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    EStreamFrame
    -0.81
    venge
    -0.70
    tro
    -0.63
    hop
    -0.63
    edo
    -0.62
     clust
    -0.61
    prime
    -0.60
     distingu
    -0.59
    erate
    -0.59
     unw
    -0.59
    POSITIVE LOGITS
     onwards
    0.73
    -'
    0.71
    1968
    0.70
    å¹
    0.69
     onward
    0.68
    1966
    0.68
     census
    0.66
     Chevrolet
    0.66
     BCE
    0.65
    1967
    0.65
    Act Density 0.049%

    No Known Activations