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

    Neuronpedia

    INDEX
    Explanations

    phrases related to potential dangers or catastrophes

    oai_token-act-pair · gpt-3.5-turbo

    references to potential disasters and crises

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     descript
    -0.77
    esse
    -0.75
     mentors
    -0.74
    ppings
    -0.74
    mates
    -0.74
     qualities
    -0.73
     pointers
    -0.73
     traits
    -0.73
     attributes
    -0.71
     instructors
    -0.70
    POSITIVE LOGITS
     catastrophe
    1.59
     collapse
    1.47
     meltdown
    1.47
     apocalypse
    1.40
     disaster
    1.36
     eruption
    1.35
     catastrophic
    1.32
     calam
    1.28
     uprising
    1.28
     rupture
    1.25
    Act Density 0.281%

    No Known Activations