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

    Neuronpedia

    INDEX
    Explanations

    data entities like dates, names, and positions

    oai_token-act-pair · gpt-3.5-turbo

    utterly blank or empty segments in text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    edIn
    -0.59
    —-
    -0.58
    inav
    -0.55
    ï
    -0.54
    dates
    -0.54
     Seym
    -0.53
    prise
    -0.51
    Make
    -0.50
    Take
    -0.50
     Ensure
    -0.48
    POSITIVE LOGITS
     located
    0.76
     hereby
    0.75
     able
    0.70
     unable
    0.70
    nt
    0.70
     incapable
    0.69
     situated
    0.69
    rael
    0.66
     supposed
    0.66
     born
    0.66
    Act Density 0.540%

    No Known Activations