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

    Neuronpedia

    INDEX
    Explanations

    comparisons with an emphasis on being the most extreme or superior

    oai_token-act-pair · gpt-3.5-turbo

    references to comparisons with "other" entities or situations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    isters
    -0.76
    onics
    -0.73
    ethe
    -0.70
    ordes
    -0.70
    ensibly
    -0.67
    ernels
    -0.66
    bows
    -0.65
    gets
    -0.64
    chairs
    -0.64
    ãĥĵ
    -0.63
    POSITIVE LOGITS
    worldly
    1.54
     aspect
    1.01
     conceivable
    1.00
    iator
    0.99
     circumstance
    0.97
     entity
    0.95
     outlet
    0.92
     imaginable
    0.91
     avenue
    0.87
     option
    0.86
    Act Density 0.055%

    No Known Activations