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

    Neuronpedia

    INDEX
    Explanations

    phrases or words denoting things that are rare, unusual, or have never been done before

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "unprecedented" and related adjectives indicating rarity or novelty

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    claimer
    -0.80
    ©¶æ
    -0.71
    belt
    -0.70
    elsen
    -0.68
    burn
    -0.67
    oras
    -0.67
    udeb
    -0.66
    adr
    -0.66
    amar
    -0.64
    =-=-=-=-
    -0.64
    POSITIVE LOGITS
     occurrence
    0.95
     occurrences
    0.88
     circumstances
    0.87
    judicial
    0.82
    LY
    0.82
    ly
    0.81
    amount
    0.76
     unanim
    0.76
    ties
    0.75
    IAL
    0.75
    Act Density 0.065%

    No Known Activations