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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating that something has not been done, seen, revealed, produced, or tracked yet

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "been" and its variants to identify discussions about experiences, actions, or states of being

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    DragonMagazine
    -0.64
    Protect
    -0.61
    lah
    -0.59
    ranging
    -0.59
    LU
    -0.59
    abba
    -0.58
    çīĪ
    -0.58
    abb
    -0.58
    Tes
    -0.58
    nee
    -0.57
    POSITIVE LOGITS
     anywhere
    1.05
     nor
    1.01
     anymore
    0.99
    yet
    0.94
     sufficiently
    0.94
     slightest
    0.93
     anything
    0.93
     yet
    0.91
     anybody
    0.88
     dime
    0.86
    Act Density 0.129%

    No Known Activations