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

    Neuronpedia

    INDEX
    Explanations

    proper nouns related to people or characters

    oai_token-act-pair · gpt-3.5-turbo

    references to female characters or entities associated with "Girl."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     constitu
    -0.86
    umbn
    -0.82
    ooming
    -0.75
     hypot
    -0.71
    uckland
    -0.69
    ascular
    -0.69
     stem
    -0.66
     centralized
    -0.66
    henko
    -0.65
    neum
    -0.65
    POSITIVE LOGITS
     Girl
    1.22
    Girl
    1.14
     Scouts
    1.13
     Girls
    1.08
     Scout
    1.02
     Jacket
    0.97
     Thing
    0.95
    Girls
    0.94
     Doll
    0.92
     Woman
    0.91
    Act Density 0.012%

    No Known Activations