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

    Neuronpedia

    INDEX
    Explanations

    phrases with a strong emotional or opinionated tone

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    ielding
    -0.62
    itatively
    -0.62
     Dayton
    -0.60
     pockets
    -0.58
    feet
    -0.57
    hips
    -0.57
    Priv
    -0.57
     Uni
    -0.57
    present
    -0.56
    nda
    -0.56
    POSITIVE LOGITS
    chy
    1.05
     ain
    0.97
    asca
    0.90
    iner
    0.89
    unes
    0.89
     happened
    0.88
     seems
    0.87
    self
    0.86
     wasn
    0.85
     begg
    0.83
    Act Density 11.707%

    No Known Activations