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

    Neuronpedia

    INDEX
    Explanations

    punctuation marks within text strings

    oai_token-act-pair · gpt-3.5-turbo

    instances of punctuation or specific grammatical structures in text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    iam
    -0.57
    ety
    -0.57
    xt
    -0.56
    estinal
    -0.54
    zel
    -0.54
    ily
    -0.54
    ¬¼
    -0.53
     Travels
    -0.53
    atars
    -0.53
    irect
    -0.53
    POSITIVE LOGITS
     nor
    1.98
     anymore
    1.39
    nor
    1.27
    yet
    1.20
     unless
    1.13
    unless
    1.06
     except
    1.01
     though
    0.92
     but
    0.92
     preferring
    0.92
    Act Density 0.413%

    No Known Activations