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

    Neuronpedia

    INDEX
    Explanations

    cultural and linguistic terms from various languages

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
     accrued
    -0.86
     increment
    -0.77
     Blair
    -0.76
     launch
    -0.76
     waiver
    -0.76
     resume
    -0.75
     resur
    -0.73
     timely
    -0.73
     retailers
    -0.72
     creep
    -0.72
    POSITIVE LOGITS
    sic
    1.87
    pron
    1.69
    meaning
    1.65
    literally
    1.64
    formerly
    1.46
    aka
    1.44
    pictured
    1.41
    sometimes
    1.34
    also
    1.33
    see
    1.31
    Act Density 0.097%

    No Known Activations