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

    Neuronpedia

    INDEX
    Explanations

    contractions in text

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating negative actions or consequences

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     Gaul
    -0.71
     Jihad
    -0.68
     Fuji
    -0.62
     Jem
    -0.61
     Juda
    -0.61
     Quart
    -0.60
     Skydragon
    -0.60
     Guards
    -0.60
     Kus
    -0.59
     camer
    -0.59
    POSITIVE LOGITS
    tarian
    0.97
    t
    0.91
    ember
    0.91
    vest
    0.91
    vable
    0.89
    ude
    0.87
    agree
    0.86
    ï¸ı
    0.83
    tre
    0.82
    sure
    0.82
    Act Density 0.159%

    No Known Activations