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

    Neuronpedia

    INDEX
    Explanations

    periods, punctuation, and abbreviations

    oai_token-act-pair · gpt-3.5-turbo

    punctuation and sentence structures in text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     occupancy
    -0.70
     retri
    -0.64
     Mechdragon
    -0.64
    unts
    -0.63
     polic
    -0.61
     enriched
    -0.61
     transporter
    -0.60
     occup
    -0.60
     occupation
    -0.59
     occupations
    -0.58
    POSITIVE LOGITS
     Specifically
    0.90
     Instead
    0.89
     Turns
    0.89
    Specifically
    0.81
    Instead
    0.80
     Previously
    0.80
     Apparently
    0.76
    arently
    0.75
     Granted
    0.74
     Honestly
    0.74
    Act Density 0.766%

    No Known Activations