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

    Neuronpedia

    INDEX
    Explanations

    connecting words indicating contrast or contradiction

    oai_token-act-pair · gpt-3.5-turbo

    the word "However" used to indicate contrast or exception

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     Goth
    -0.68
    board
    -0.62
     award
    -0.59
     Mountains
    -0.58
    stein
    -0.57
     triangle
    -0.57
    advertising
    -0.56
     parade
    -0.56
    screen
    -0.55
    letter
    -0.55
    POSITIVE LOGITS
    tons
    0.92
    chers
    0.80
    lers
    0.79
    chery
    0.76
     interestingly
    0.73
    ler
    0.72
     importantly
    0.70
    herty
    0.67
    LER
    0.67
    tif
    0.66
    Act Density 0.030%

    No Known Activations