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

    Neuronpedia

    INDEX
    Explanations

    superlatives or comparisons, such as "best of the best" or "weakest of the trilogy"

    oai_token-act-pair · gpt-3.5-turbo

    phrases denoting rankings or lists of the best or worst entities

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    BER
    -0.72
    ARA
    -0.70
    RANT
    -0.69
    lier
    -0.69
    ova
    -0.66
    zon
    -0.63
    zer
    -0.61
    imaru
    -0.61
    isson
    -0.61
    EMA
    -0.60
    POSITIVE LOGITS
     bunch
    0.98
     proverbial
    0.84
     millennium
    0.79
     latter
    0.79
     trio
    0.79
     pack
    0.75
     worst
    0.74
     world
    0.74
     litter
    0.74
     largest
    0.72
    Act Density 0.201%

    No Known Activations