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

    Neuronpedia

    INDEX
    Explanations

    instances of quantities being more than a specific number

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating a quantity or numerical count

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    vantage
    -0.80
    xtap
    -0.80
    }}}
    -0.79
    essage
    -0.78
    ĺħ
    -0.71
    hyde
    -0.71
    atan
    -0.70
    reciation
    -0.68
    liest
    -0.66
    rette
    -0.66
    POSITIVE LOGITS
     than
    2.10
    than
    1.51
     Than
    1.46
     stringent
    0.77
     importantly
    0.72
     mouths
    0.71
     then
    0.69
     TH
    0.69
     modest
    0.69
     bang
    0.67
    Act Density 0.106%

    No Known Activations