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

    Neuronpedia

    INDEX
    Explanations

    terms related to hazards, especially workplace hazards

    oai_token-act-pair · gpt-3.5-turbo

    references to various types of hazards

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ebus
    -0.96
    gdala
    -0.95
    ulet
    -0.89
    itton
    -0.87
    artney
    -0.85
    ivia
    -0.84
    het
    -0.84
    istors
    -0.78
    hetti
    -0.74
    ampires
    -0.74
    POSITIVE LOGITS
     hazards
    1.12
     hazard
    1.09
    ously
    0.97
     Hazard
    0.89
     mitigation
    0.82
    sofar
    0.71
    ous
    0.71
     Bunny
    0.69
     unbeliev
    0.68
    saf
    0.68
    Act Density 0.011%

    No Known Activations