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

    Neuronpedia

    INDEX
    Explanations

    instances where the word "one" is being emphasized

    oai_token-act-pair · gpt-3.5-turbo

    references to absence or lack of quantity

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    yrinth
    -0.68
    alth
    -0.66
    types
    -0.63
    bos
    -0.63
     notor
    -0.62
    atu
    -0.61
     sorts
    -0.60
    omsday
    -0.59
    comings
    -0.59
    loo
    -0.59
    POSITIVE LOGITS
     ounce
    1.20
     dime
    1.20
     penny
    0.98
     shred
    0.88
     nor
    0.88
     clue
    0.84
     inch
    0.82
     slightest
    0.80
     whatsoever
    0.78
     doubt
    0.78
    Act Density 0.121%

    No Known Activations