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

    Neuronpedia

    INDEX
    Explanations

    conjunctions, specifically the word "but"

    oai_token-act-pair · gpt-3.5-turbo

    the word "but" indicating contrast or contradiction

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    agra
    -0.79
    tnc
    -0.72
    itto
    -0.67
    idon
    -0.64
    entry
    -0.62
    analysis
    -0.62
    代
    -0.62
    itaire
    -0.61
    olution
    -0.61
    venue
    -0.61
    POSITIVE LOGITS
    tons
    1.10
    chery
    0.83
    chers
    0.75
    ts
    0.74
     nevertheless
    0.72
    still
    0.69
     alas
    0.69
     nonetheless
    0.68
     hey
    0.64
    tern
    0.63
    Act Density 0.107%

    No Known Activations