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

    Neuronpedia

    INDEX
    Explanations

    phrases relating to comparisons and contrasts, specifically emphasizing the uniqueness of something

    oai_token-act-pair · gpt-3.5-turbo

    comparative phrases highlighting uniqueness or absence of alternatives

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    abin
    -0.84
     Reloaded
    -0.67
    AFTA
    -0.66
    fa
    -0.63
    zai
    -0.63
     Different
    -0.62
    uler
    -0.62
    ãĤ©
    -0.61
    uddin
    -0.58
    antics
    -0.58
    POSITIVE LOGITS
    worldly
    1.09
     nor
    0.92
     dime
    0.85
     option
    0.83
    nor
    0.80
     practicable
    0.73
     occurrences
    0.73
     whatsoever
    0.70
    icable
    0.70
     surprises
    0.68
    Act Density 0.050%

    No Known Activations