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

    Neuronpedia

    INDEX
    Explanations

    words associated with challenging authority or societal norms

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "dare" and its variations, often indicating challenges or confrontations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    urgy
    -0.74
    ulator
    -0.70
    ulators
    -0.69
    effic
    -0.68
     Rite
    -0.67
    iple
    -0.65
    utra
    -0.65
    ulatory
    -0.64
    OTOS
    -0.62
    Tool
    -0.62
    POSITIVE LOGITS
     dare
    1.02
     Dare
    0.96
     daring
    0.89
     defy
    0.86
    ngth
    0.83
    evil
    0.83
     dared
    0.81
     roam
    0.80
     boldly
    0.75
     provoke
    0.72
    Act Density 0.026%

    No Known Activations