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

    Neuronpedia

    INDEX
    Explanations

    contractions of words (e.g., "shouldn't")

    oai_token-act-pair · gpt-3.5-turbo

    instances of the phrase "shouldn't," indicating a focus on caution or restraint

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     locating
    -0.62
     wheelchair
    -0.61
     indexed
    -0.58
     Resurrection
    -0.58
     seeded
    -0.57
     Letter
    -0.57
     mailing
    -0.56
     offline
    -0.56
     Published
    -0.56
     fulfilled
    -0.55
    POSITIVE LOGITS
    't
    1.41
    n
    1.12
    ighed
    0.91
    ny
    0.88
    ÃŃ
    0.87
    eness
    0.87
    ned
    0.87
    nor
    0.85
    ada
    0.83
    uts
    0.83
    Act Density 0.016%

    No Known Activations