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

    Neuronpedia

    INDEX
    Explanations

    specific numeric patterns

    oai_token-act-pair · gpt-3.5-turbo

    numerical values associated with specific percentages and quantities

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    eworks
    -0.89
    istically
    -0.72
    boarding
    -0.68
    matically
    -0.65
     immersion
    -0.64
    lining
    -0.62
    aley
    -0.62
     clad
    -0.61
    ushes
    -0.61
    igating
    -0.61
    POSITIVE LOGITS
    999
    1.37
    LECT
    1.01
    089
    0.87
    059
    0.83
    CRIP
    0.81
    9999
    0.81
    onso
    0.79
    090
    0.79
    ¢
    0.79
    IELD
    0.79
    Act Density 0.035%

    No Known Activations