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

    Neuronpedia

    INDEX
    Explanations

    words related to counterpoints or contrasting views followed by a strong stance or argument

    oai_token-act-pair · gpt-3.5-turbo

    the end of text tokens

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    agra
    -0.76
    itto
    -0.69
    kie
    -0.64
    oufl
    -0.64
    chnology
    -0.60
    analysis
    -0.60
     vein
    -0.59
    idon
    -0.59
    tnc
    -0.59
    rongh
    -0.58
    POSITIVE LOGITS
    tons
    1.09
     nevertheless
    0.88
     alas
    0.87
     nonetheless
    0.87
    chery
    0.84
     fortunately
    0.79
    chers
    0.74
     luckily
    0.73
     hey
    0.69
     GHC
    0.68
    Act Density 0.099%

    No Known Activations