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

    Neuronpedia

    INDEX
    Explanations

    phrases or sentences that are marked in the text with a specific separator (Ċ)

    oai_token-act-pair · gpt-3.5-turbo

    general patterns of numerical data or statistical information

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    heit
    -0.77
    boro
    -0.72
    ounce
    -0.62
    abouts
    -0.61
    utonium
    -0.61
    escape
    -0.60
    hement
    -0.60
    uto
    -0.58
    utic
    -0.57
    ocial
    -0.57
    POSITIVE LOGITS
    ³³³
    0.75
    reditary
    0.71
    ³³³³
    0.66
    ³³³³³³³³
    0.64
    Sensor
    0.62
    However
    0.61
    ³³³³³³³³³³³³³³³³
    0.61
    Lear
    0.61
    Scroll
    0.61
    ccording
    0.60
    Act Density 0.746%

    No Known Activations