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

    Neuronpedia

    INDEX
    Explanations

    phrases containing conversational fillers like "Okay" and "so"

    oai_token-act-pair · gpt-3.5-turbo

    conversational elements and discourse markers in dialogues

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    dayName
    -0.84
    resso
    -0.78
    reated
    -0.76
    conservancy
    -0.75
    è¦ļéĨĴ
    -0.72
    machine
    -0.70
    string
    -0.68
    à¨
    -0.68
    lain
    -0.68
    nai
    -0.67
    POSITIVE LOGITS
     bye
    0.94
     congr
    0.82
     Wrong
    0.81
    bye
    0.80
     caveats
    0.78
     congratulations
    0.77
     maybe
    0.73
     yeah
    0.71
     caveat
    0.70
     sorry
    0.69
    Act Density 0.055%

    No Known Activations