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

    Neuronpedia

    INDEX
    Explanations

    contractions and possessive pronouns followed by a verb

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating certainty or affirmation

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     Goldberg
    -0.65
     Constitutional
    -0.61
     supra
    -0.59
     Fiscal
    -0.58
     Coffin
    -0.57
    asca
    -0.57
     Morocco
    -0.57
    ccording
    -0.55
    milo
    -0.55
     Triple
    -0.54
    POSITIVE LOGITS
     accordingly
    0.70
    forth
    0.65
    hung
    0.65
    eming
    0.65
    emis
    0.64
    imaru
    0.63
     LIA
    0.60
    instead
    0.60
    phys
    0.59
    apter
    0.59
    Act Density 0.347%

    No Known Activations