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

    Neuronpedia

    INDEX
    Explanations

    phrases related to identifying and discussing risks

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to risks and potential dangers

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ergy
    -0.94
    gdala
    -0.78
    cle
    -0.78
    gran
    -0.76
    Hat
    -0.71
    Nap
    -0.71
    eve
    -0.70
    eenth
    -0.70
     Kinnikuman
    -0.69
    MT
    -0.69
    POSITIVE LOGITS
     risks
    1.09
     risk
    0.93
     dangers
    0.88
     pitfalls
    0.88
    afety
    0.87
     hazards
    0.87
     endanger
    0.84
    crow
    0.78
     jeopard
    0.77
     consequences
    0.77
    Act Density 0.013%

    No Known Activations