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

    Neuronpedia

    INDEX
    Explanations

    the word "Source" followed by a number

    oai_token-act-pair · gpt-3.5-turbo

    references to "Source" or citations within the text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    akuya
    -0.69
     haste
    -0.67
     benches
    -0.64
     staggered
    -0.63
    cases
    -0.63
     atro
    -0.63
     specials
    -0.62
    oooooooo
    -0.61
    onom
    -0.61
     gall
    -0.61
    POSITIVE LOGITS
     Source
    3.93
    Source
    2.61
     Sources
    2.15
     source
    1.96
    source
    1.92
    Sources
    1.72
     sources
    1.51
    SOURCE
    1.49
    OURCE
    1.46
    ources
    1.39
    Act Density 0.006%

    No Known Activations