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

    Neuronpedia

    INDEX
    Explanations

    updates or revisions within a text

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "UPDATE."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    atom
    -0.83
    azes
    -0.81
    ung
    -0.78
    odor
    -0.77
    ü
    -0.77
    fle
    -0.77
    rift
    -0.76
    course
    -0.75
    uana
    -0.74
    stood
    -0.74
    POSITIVE LOGITS
     UPDATE
    1.25
    UPDATE
    1.17
    ION
    1.09
    CLAIM
    1.06
    BOX
    1.02
     EDIT
    1.02
    INGTON
    0.99
     ABOUT
    0.97
    BOOK
    0.97
     REPORT
    0.97
    Act Density 0.011%

    No Known Activations