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

    Neuronpedia

    INDEX
    Explanations

    email-related features like invalid address and robot verification prompts

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    ciation
    -0.50
    ciples
    -0.49
    gow
    -0.44
     Vengeance
    -0.42
     ranges
    -0.41
    seys
    -0.40
     Formation
    -0.40
    multipl
    -0.40
    semb
    -0.39
    urst
    -0.39
    POSITIVE LOGITS
     anymore
    0.60
    iframe
    0.53
    à¼
    0.50
    taboola
    0.50
     externalToEVAOnly
    0.44
     subscribing
    0.44
    ATED
    0.43
    ======
    0.43
     isEnabled
    0.41
     nor
    0.41
    Act Density 12.506%

    No Known Activations