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

    Neuronpedia

    INDEX
    Explanations

    conversational phrases starting with "Well,"

    oai_token-act-pair · gpt-3.5-turbo

    phrases and transitions that indicate conversational flow or dialogue

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ealous
    -0.77
    Winged
    -0.74
    ebted
    -0.71
    keyes
    -0.71
    âĵĺ
    -0.71
    è»
    -0.69
    pering
    -0.67
    Higher
    -0.65
    mone
    -0.64
    foreseen
    -0.64
    POSITIVE LOGITS
     yeah
    1.13
     guess
    1.10
     luckily
    1.05
     maybe
    1.04
     fortunately
    1.02
     uh
    1.02
     hello
    1.02
     congr
    0.95
     technically
    0.94
     okay
    0.94
    Act Density 0.072%

    No Known Activations