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

    Neuronpedia

    INDEX
    Explanations

    instances related to physical violence and abuse

    oai_token-act-pair · gpt-3.5-turbo

    terms associated with violence and abuse

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    soDeliveryDate
    -0.96
    soType
    -0.92
    isSpecialOrderable
    -0.85
    izen
    -0.77
    alg
    -0.73
    aer
    -0.72
    agall
    -0.72
    natureconservancy
    -0.71
    travel
    -0.70
     Balance
    -0.69
    POSITIVE LOGITS
     abusive
    0.92
     unarmed
    0.88
     victims
    0.83
     handcuffed
    0.80
     abusers
    0.79
     interrog
    0.78
     punishments
    0.78
     violence
    0.78
     violently
    0.77
     pepper
    0.77
    Act Density 0.193%

    No Known Activations