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

    Neuronpedia

    INDEX
    Explanations

    quantitative measurements related to technology, such as memory capacity and encryption bit rates

    oai_token-act-pair · gpt-3.5-turbo

    numerical values related to data storage capacities and encryption strengths

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    orie
    -0.78
    ansen
    -0.76
    andise
    -0.76
    orical
    -0.75
     Nau
    -0.74
    agne
    -0.71
    raising
    -0.66
    men
    -0.65
    oric
    -0.65
    avis
    -0.65
    POSITIVE LOGITS
    256
    0.93
    384
    0.90
     CTR
    0.87
    1024
    0.83
     bytes
    0.82
    GB
    0.78
     integers
    0.77
    GBT
    0.74
    ãĥ¢
    0.74
    Minecraft
    0.74
    Act Density 0.020%

    No Known Activations