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

    Neuronpedia

    INDEX
    Explanations

    phrases where someone expresses their desire or intent to perform an action

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    requires
    -0.71
    soever
    -0.66
    quartered
    -0.63
    currently
    -0.58
    Recommended
    -0.58
    soDeliveryDate
    -0.56
     presently
    -0.56
    interstitial
    -0.56
     Appears
    -0.56
    lies
    -0.56
    POSITIVE LOGITS
     postpone
    0.85
     capitalize
    0.84
     evacuate
    0.81
     eliminate
    0.77
     accomplish
    0.77
     reassure
    0.76
     insure
    0.76
     awaken
    0.74
     develop
    0.74
     introduce
    0.74
    Act Density 13.466%

    No Known Activations