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

    Neuronpedia

    INDEX
    Explanations

    references to magical or supernatural threats

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     Downtown
    -0.18
     Vietnam
    -0.17
     Mexican
    -0.17
     Vietnamese
    -0.16
     Pepsi
    -0.16
     Pulitzer
    -0.16
     American
    -0.15
     Hoover
    -0.15
    American
    -0.15
     Indonesian
    -0.15
    POSITIVE LOGITS
     mage
    0.24
     Clan
    0.21
     orc
    0.21
     Elves
    0.20
     elves
    0.19
     warriors
    0.19
     Gry
    0.19
     Gor
    0.19
     wizard
    0.19
     swords
    0.18
    Act Density 0.385%

    No Known Activations