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

    Neuronpedia

    INDEX
    Explanations

    references to software bug fixes or patches

    oai_token-act-pair · gpt-3.5-turbo

    references to specific bugs and fixes in software updates

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    tainment
    -0.74
    reviewed
    -0.68
    forts
    -0.67
    Äĵ
    -0.67
    ®,
    -0.67
    —-
    -0.66
    âĢ¦âĢ¦âĢ¦âĢ¦
    -0.66
     todd
    -0.65
     juven
    -0.64
    frog
    -0.64
    POSITIVE LOGITS
     incorrectly
    1.31
     incorrect
    1.19
     errone
    1.16
     tooltip
    1.14
     improperly
    1.12
     wrongly
    1.09
     typo
    1.06
     crash
    1.04
     erroneous
    0.97
     sometimes
    0.97
    Act Density 0.184%

    No Known Activations