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

    Neuronpedia

    INDEX
    Explanations

    instances of the word "very"

    oai_token-act-pair · gpt-3.5-turbo

    intensifiers that emphasize feelings or states of being

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    adelphia
    -0.78
    sburgh
    -0.74
    ourses
    -0.71
    olor
    -0.68
    onis
    -0.68
    aneers
    -0.65
    osal
    -0.65
    icago
    -0.64
    ses
    -0.63
    osis
    -0.63
    POSITIVE LOGITS
     rare
    0.88
     important
    0.84
     informative
    0.84
     interesting
    0.82
     different
    0.82
     unusual
    0.81
     unlikely
    0.80
     difficult
    0.80
     exciting
    0.79
     secretive
    0.77
    Act Density 0.065%

    No Known Activations