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

    Neuronpedia

    INDEX
    Explanations

    phrases related to surprises

    oai_token-act-pair · gpt-3.5-turbo

    phrasing related to expectations or surprises

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    audi
    -0.66
    abulary
    -0.62
    erential
    -0.61
    ortex
    -0.59
     filibuster
    -0.57
    oyer
    -0.57
    orce
    -0.56
     Suc
    -0.56
     Logged
    -0.56
    adelphia
    -0.55
    POSITIVE LOGITS
    pires
    0.81
     follows
    0.74
    pired
    0.73
     regards
    0.68
     unwelcome
    0.65
     well
    0.64
    lies
    0.63
     soon
    0.61
    piration
    0.61
     opposed
    0.60
    Act Density 0.043%

    No Known Activations