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

    Neuronpedia

    INDEX
    Explanations

    contrasting elements or concepts

    oai_token-act-pair · gpt-3.5-turbo

    phrases or clauses that contrast or negate a preceding statement

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ct
    -0.75
    enter
    -0.72
    nat
    -0.69
    ory
    -0.67
    asons
    -0.66
    uter
    -0.65
    uther
    -0.65
    wake
    -0.65
    tty
    -0.61
    ctor
    -0.61
    POSITIVE LOGITS
     nevertheless
    1.03
     nonetheless
    0.95
     rather
    0.95
     luckily
    0.92
     fortunately
    0.92
     suffice
    0.87
     alas
    0.83
     hey
    0.81
    rather
    0.81
     thankfully
    0.79
    Act Density 0.130%

    No Known Activations