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

    Neuronpedia

    INDEX
    Explanations

    words related to documents and data analysis

    oai_token-act-pair · gpt-3.5-turbo

    phrases that indicate ownership or possession

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     lately
    -0.62
    ideos
    -0.57
    traditional
    -0.55
    isal
    -0.55
     relies
    -0.54
     terribly
    -0.54
    orks
    -0.53
    journal
    -0.53
     struggles
    -0.53
    andals
    -0.53
    POSITIVE LOGITS
     whichever
    1.52
     each
    1.39
     every
    1.23
     whoever
    1.18
     EVERY
    1.08
    each
    1.01
     all
    1.00
     ALL
    0.97
     any
    0.96
     whatever
    0.93
    Act Density 0.568%

    No Known Activations