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

    Neuronpedia

    INDEX
    Explanations

    proper nouns related to geopolitical events or political figures

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    cember
    -1.10
    inated
    -0.94
    istries
    -0.92
    isphere
    -0.91
    aukee
    -0.91
    eering
    -0.90
    bush
    -0.90
    apult
    -0.89
    boarding
    -0.88
    acted
    -0.86
    POSITIVE LOGITS
     anyone
    0.98
     anybody
    0.94
    actic
    0.92
     corrections
    0.85
     dozens
    0.82
     Greenberg
    0.82
     many
    0.81
     Scarlet
    0.80
     usual
    0.80
    iability
    0.80
    Act Density 0.285%

    No Known Activations