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

    Neuronpedia

    INDEX
    Explanations

    names of people or characters followed by a number or symbol

    oai_token-act-pair · gpt-3.5-turbo

    specific numerical values and accompanying symbols

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     millions
    -0.67
     thousands
    -0.66
     snowball
    -0.65
     tens
    -0.64
     billions
    -0.61
     hundreds
    -0.61
     harness
    -0.56
     mush
    -0.55
     trillions
    -0.54
    eez
    -0.54
    POSITIVE LOGITS
     respectively
    1.03
     Sakuya
    0.81
     CrossRef
    0.78
    +.
    0.77
    .).
    0.76
     Lastly
    0.74
    )).
    0.73
    ));
    0.73
    ]).
    0.71
    ).
    0.69
    Act Density 0.334%

    No Known Activations