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

    Neuronpedia

    INDEX
    Explanations

    numeric data, such as page numbers, amounts of money, and dates

    oai_token-act-pair · gpt-3.5-turbo

    brackets or bracket-like symbols in the text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     cradle
    -0.72
    userc
    -0.71
    ãĤ©
    -0.68
    ĵĺ
    -0.68
    boro
    -0.66
    ury
    -0.65
    acies
    -0.65
    oe
    -0.63
    rette
    -0.62
     slam
    -0.62
    POSITIVE LOGITS
    ...]
    0.96
    oldemort
    0.86
     However
    0.80
    amazon
    0.79
     Conversely
    0.77
    TPS
    0.76
    ][
    0.75
     Flavoring
    0.73
     Additionally
    0.73
    Attempts
    0.72
    Act Density 0.045%

    No Known Activations