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

    Neuronpedia

    INDEX
    Explanations

    phrases related to criminal activities and intentions

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to the act of committing various actions, particularly negative ones

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     complexion
    -0.71
    sama
    -0.67
    atin
    -0.67
    nas
    -0.65
    ****************
    -0.64
    addin
    -0.64
    dayName
    -0.63
    adish
    -0.63
     allergies
    -0.61
    abwe
    -0.61
    POSITIVE LOGITS
     suicide
    0.91
     committed
    0.87
     adultery
    0.85
     committing
    0.84
    utions
    0.83
    acts
    0.80
    itent
    0.77
     commit
    0.75
    ruction
    0.75
    venant
    0.74
    Act Density 0.024%

    No Known Activations