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

    Neuronpedia

    INDEX
    Explanations

    contractions specifically with "I am" written as "I'm"

    oai_token-act-pair · gpt-3.5-turbo

    instances of self-identification or personal statements

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ainer
    -0.70
     mater
    -0.69
     separates
    -0.64
     hinder
    -0.62
     looms
    -0.61
     occurs
    -0.61
     perish
    -0.61
     membr
    -0.60
     entails
    -0.60
     Tactics
    -0.59
    POSITIVE LOGITS
     gonna
    1.27
     glad
    0.95
     thankful
    0.93
     fortunate
    0.92
     lucky
    0.91
     hoping
    0.90
     going
    0.89
     gotta
    0.87
     sorry
    0.87
     afraid
    0.86
    Act Density 0.034%

    No Known Activations