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

    Neuronpedia

    INDEX
    Explanations

    parts of the text with emphasis markers

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "emphasis" and related phrasing

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    tre
    -0.72
    izons
    -0.71
     Rebell
    -0.68
    STON
    -0.68
    wagen
    -0.68
    tein
    -0.67
    met
    -0.66
     Skydragon
    -0.66
    Nazis
    -0.65
    apons
    -0.63
    POSITIVE LOGITS
     emphasis
    1.05
    phasis
    0.93
    xual
    0.78
     shifts
    0.78
    inction
    0.76
     reliance
    0.76
    estinal
    0.75
    uality
    0.75
    baugh
    0.73
    erences
    0.73
    Act Density 0.018%

    No Known Activations