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

    Neuronpedia

    INDEX
    Explanations

    phrases related to completion or entirety, often including the word 'full' preceding a number or a noun

    oai_token-act-pair · gpt-3.5-turbo

    phrases that include the word "full" indicating comprehensive elements or entirety

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ppings
    -0.65
     Tycoon
    -0.65
    ghost
    -0.65
    artifacts
    -0.65
    inventoryQuantity
    -0.64
     chants
    -0.64
    pher
    -0.64
    lees
    -0.64
    ðŁĺ
    -0.63
     Guys
    -0.63
    POSITIVE LOGITS
     blown
    1.02
    erton
    0.96
     fled
    0.94
     complement
    0.92
    hearted
    0.87
     sized
    0.85
    fledged
    0.85
     frontal
    0.83
    eding
    0.80
    ening
    0.80
    Act Density 0.026%

    No Known Activations