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

    Neuronpedia

    INDEX
    Explanations

    phrases emphasizing a particular characteristic with the word "well" preceding it

    oai_token-act-pair · gpt-3.5-turbo

    the phrase "well" indicating opinions or evaluations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    20439
    -0.89
    ascus
    -0.78
    osponsors
    -0.76
    yk
    -0.76
    actionDate
    -0.73
    itive
    -0.72
    æĪ¦
    -0.71
    ription
    -0.69
    gradient
    -0.66
    rament
    -0.65
    POSITIVE LOGITS
     dunno
    0.81
     bye
    0.70
     yeah
    0.68
    yeah
    0.63
     guessed
    0.63
    imately
    0.62
    mmm
    0.62
    entimes
    0.61
    ..
    0.61
    ...
    0.61
    Act Density 0.065%

    No Known Activations