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

    Neuronpedia

    INDEX
    Explanations

    word sequences that indicate future actions or events

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    lees
    -0.86
    xual
    -0.74
    lee
    -0.71
     Franch
    -0.70
    then
    -0.70
    theless
    -0.69
    olor
    -0.67
    ords
    -0.67
    still
    -0.66
    tics
    -0.66
    POSITIVE LOGITS
     generation
    1.21
     installment
    1.10
     step
    1.03
     iteration
    0.99
     phase
    0.98
     logical
    0.98
     batch
    0.95
     millenn
    0.93
     decade
    0.92
     chapter
    0.89
    Act Density 0.048%

    No Known Activations