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

    Neuronpedia

    INDEX
    Explanations

    pairs of words hyphenated together in a specific pattern

    oai_token-act-pair · gpt-3.5-turbo

    mentions of procedural processes or steps related to rules and regulations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     aborted
    -0.61
     migrated
    -0.60
     rubbed
    -0.60
     persisted
    -0.60
     Gork
    -0.60
     Fusion
    -0.60
     stoked
    -0.59
     killed
    -0.59
    ioned
    -0.58
     revived
    -0.57
    POSITIVE LOGITS
    committee
    0.95
    elect
    0.88
    election
    0.84
    step
    0.83
    command
    0.83
    product
    0.82
    products
    0.82
    tem
    0.81
    default
    0.80
    book
    0.79
    Act Density 0.033%

    No Known Activations