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

    Neuronpedia

    INDEX
    Explanations

    instances of "first-ever" events or descriptions

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    gre
    -0.63
    icter
    -0.58
    nah
    -0.57
    yards
    -0.57
    elo
    -0.56
    masters
    -0.55
    rats
    -0.54
    rage
    -0.54
    ole
    -0.53
    olith
    -0.53
    POSITIVE LOGITS
     foray
    0.68
     edition
    0.65
     iteration
    0.64
     rebound
    0.61
     glimpse
    0.61
     installment
    0.56
     inaugural
    0.56
     occurrence
    0.56
     global
    0.55
     attempt
    0.54
    Act Density 9.689%

    No Known Activations