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

    Neuronpedia

    INDEX
    Explanations

    comparisons indicating a preference or choice

    oai_token-act-pair · gpt-3.5-turbo

    instances where the word "rather" is used to introduce contrasting statements or opinions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     unknown
    -0.66
     margin
    -0.59
     extinct
    -0.59
     finals
    -0.58
     extension
    -0.58
     heights
    -0.56
     hanging
    -0.56
     margins
    -0.55
     quarters
    -0.55
     playoff
    -0.55
    POSITIVE LOGITS
     Rather
    3.41
    Rather
    2.59
     Instead
    2.38
    rather
    2.15
    Instead
    2.02
    instead
    1.64
     rather
    1.56
     Quite
    1.43
     Nonetheless
    1.32
     Specifically
    1.27
    Act Density 0.011%

    No Known Activations