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

    Neuronpedia

    INDEX
    Explanations

    phrases or descriptions that emphasize the importance or prominence of specific elements

    oai_token-act-pair · gpt-3.5-turbo

    phrases emphasizing characteristics or fundamental qualities of subjects

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     equivalents
    -0.67
    psons
    -0.64
    anse
    -0.63
     Unsure
    -0.62
    reys
    -0.61
    >]
    -0.61
    ositories
    -0.61
    estyles
    -0.61
    ptions
    -0.60
    uci
    -0.60
    POSITIVE LOGITS
     sorts
    0.95
     ours
    0.88
     contention
    0.78
     hers
    0.76
     mine
    0.71
     theirs
    0.69
     earners
    0.65
    Initialized
    0.64
     importance
    0.63
     Nanto
    0.60
    Act Density 0.180%

    No Known Activations