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

    Neuronpedia

    INDEX
    Explanations

    superlatives or extremes in a comparison

    oai_token-act-pair · gpt-3.5-turbo

    terms that indicate extremes or superlatives of various conditions or qualities

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     GOD
    -0.68
    enium
    -0.67
    sterdam
    -0.67
    erate
    -0.66
     heartbeat
    -0.66
    kers
    -0.65
    bara
    -0.65
    agall
    -0.63
    fect
    -0.61
    geon
    -0.61
    POSITIVE LOGITS
    ãĥ´
    0.95
    mos
    0.88
    ï¸ı
    0.87
    EStream
    0.82
    ãĥĥ
    0.75
    !/
    0.75
    MET
    0.73
    ãĤ´ãĥ³
    0.72
     sqor
    0.71
    é¾įå¥ij士
    0.69
    Act Density 0.065%

    No Known Activations