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

    Neuronpedia

    INDEX
    Explanations

    second person pronouns followed by verbs indicating desire or preference

    oai_token-act-pair · gpt-3.5-turbo

    phrases addressing the reader directly, particularly involving the word "you" and expressions of desire or requirement

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     Paraly
    -0.64
     Forth
    -0.62
    oad
    -0.61
    OX
    -0.60
    wikipedia
    -0.59
    Quotes
    -0.59
     Rusty
    -0.58
     Ike
    -0.58
    odore
    -0.58
     BMC
    -0.58
    POSITIVE LOGITS
    ammed
    0.72
     slightest
    0.71
     regul
    0.70
     ever
    0.69
     legitimately
    0.67
     weren
    0.67
     somehow
    0.66
     dared
    0.66
     sacrific
    0.65
     encount
    0.65
    Act Density 0.298%

    No Known Activations