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

    Neuronpedia

    INDEX
    Explanations

    phrases related to actions of restoration or reversal

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "restore" and related terms indicating a return to a previous state or condition

    oai_token-act-pair · gpt-3.5-turboTriggered by @jordan
    New Auto-Interp
    Negative Logits
    rics
    -0.93
    QB
    -0.85
    chens
    -0.84
    kers
    -0.82
    Cola
    -0.80
    gerald
    -0.78
    onne
    -0.75
    edin
    -0.74
    ks
    -0.73
    wer
    -0.73
    POSITIVE LOGITS
     havoc
    0.83
    itial
    0.82
     restore
    0.82
    shire
    0.81
     vig
    0.78
    imar
    0.76
     harmony
    0.76
     restoration
    0.76
     restoring
    0.74
    owship
    0.73
    Act Density 0.791%

    No Known Activations