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

    Neuronpedia

    INDEX
    Explanations

    locations where people are from

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating geographic origin or location

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ratulations
    -0.81
    faced
    -0.80
    few
    -0.78
    potion
    -0.68
    tarians
    -0.67
    onential
    -0.66
    adata
    -0.66
    attribute
    -0.65
    icipated
    -0.65
    comfort
    -0.64
    POSITIVE LOGITS
     afar
    1.44
     abroad
    1.12
     Ethiopia
    1.00
     whence
    0.98
     Latvia
    0.97
     somewhere
    0.97
     Denmark
    0.97
     Nigeria
    0.96
     Finland
    0.95
     Hungary
    0.95
    Act Density 0.105%

    No Known Activations