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

    Neuronpedia

    INDEX
    Explanations

    words related to hazards or hazards themselves

    oai_token-act-pair · gpt-3.5-turbo

    references to hazards and hazardous materials

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    gdala
    -0.96
    itton
    -0.92
    hetti
    -0.88
    ampires
    -0.87
    AIN
    -0.84
    inyl
    -0.83
    ombo
    -0.80
    orative
    -0.79
    oaded
    -0.78
    thood
    -0.77
    POSITIVE LOGITS
     hazards
    0.94
     hazard
    0.92
    ously
    0.88
     mitigation
    0.73
     Patrol
    0.73
     Hazard
    0.72
    hooting
    0.71
    SEE
    0.70
     evacuation
    0.69
     Bunny
    0.67
    Act Density 0.043%

    No Known Activations