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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating a possibility or likelihood

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    culosis
    -0.73
    anwhile
    -0.70
    rones
    -0.68
    ulously
    -0.67
    shall
    -0.66
     Mund
    -0.64
    ãĥ¼ãĥ«
    -0.63
    enaries
    -0.61
    inav
    -0.61
     guarantee
    -0.60
    POSITIVE LOGITS
     able
    1.09
     tempted
    1.05
     forgiven
    0.94
     considered
    0.94
     swayed
    0.91
     construed
    0.90
     mistaken
    0.88
     regarded
    0.88
    acons
    0.83
     viewed
    0.82
    Act Density 0.108%

    No Known Activations