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

    Neuronpedia

    INDEX
    Explanations

    instances where something is uncommon or rare

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating rarity or unusualness

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    arching
    -0.73
    iyah
    -0.70
    anooga
    -0.64
     Awakens
    -0.62
     Okay
    -0.61
    ghazi
    -0.61
     Bern
    -0.61
     BET
    -0.61
    deck
    -0.60
    inders
    -0.59
    POSITIVE LOGITS
     occurrences
    0.90
    rences
    0.89
     occurrence
    0.88
     occasions
    0.80
     rare
    0.76
    pmwiki
    0.73
     gatherings
    0.71
     imaginable
    0.71
    ãĥīãĥ©ãĤ´ãĥ³
    0.68
    ibrary
    0.68
    Act Density 0.172%

    No Known Activations