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

    Neuronpedia

    INDEX
    Explanations

    numbers followed by "year-old" as a pattern

    oai_token-act-pair · gpt-3.5-turbo

    references to the age of children, particularly two to three-year-olds

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     thr
    -0.67
     redund
    -0.64
     GOODMAN
    -0.62
     reflections
    -0.59
     Rez
    -0.58
     entropy
    -0.57
     Leth
    -0.57
     spelled
    -0.56
     tens
    -0.56
     grav
    -0.56
    POSITIVE LOGITS
    olds
    1.32
    old
    1.30
    long
    1.08
    plus
    1.05
    appointed
    1.02
    ago
    1.00
    term
    1.00
    endment
    0.99
    OLD
    0.94
    period
    0.94
    Act Density 0.031%

    No Known Activations