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

    Neuronpedia

    INDEX
    Explanations

    questions involving decision-making and moral dilemmas

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "should," indicating ethical or prescriptive considerations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    esters
    -0.64
     pie
    -0.61
     elusive
    -0.60
     witness
    -0.59
     hosp
    -0.58
     anonymously
    -0.58
     fraternity
    -0.58
     trace
    -0.56
     mystery
    -0.56
     investigators
    -0.56
    POSITIVE LOGITS
     Should
    3.44
    Should
    2.83
    should
    2.28
     SHOULD
    2.09
     should
    1.44
     Would
    1.41
     Might
    1.39
     Could
    1.38
     Shall
    1.36
     Wouldn
    1.32
    Act Density 0.013%

    No Known Activations