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

    Neuronpedia

    INDEX
    Explanations

    personal pronouns and verbs related to one's own actions

    oai_token-act-pair · gpt-3.5-turbo

    references to a specific male individual or pronoun

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    htaking
    -0.69
    earch
    -0.69
    endif
    -0.67
    anking
    -0.66
    endment
    -0.65
    history
    -0.63
    aphael
    -0.62
    ifference
    -0.61
    maxwell
    -0.61
     Bundes
    -0.61
    POSITIVE LOGITS
    'd
    1.05
     undertook
    0.94
    eded
    0.94
     redes
    0.94
    uristic
    0.92
    aped
    0.91
     wrote
    0.90
     considers
    0.90
     swore
    0.89
     dreamed
    0.88
    Act Density 0.106%

    No Known Activations