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

    Neuronpedia

    INDEX
    Explanations

    introductory phrases that transition to a new topic or point in the text

    oai_token-act-pair · gpt-3.5-turbo

    transitional phrases that indicate a change in subject or introduce new topics

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     vandal
    -0.65
     mull
    -0.64
     matured
    -0.60
     flame
    -0.59
     revered
    -0.59
     Maid
    -0.59
     settlements
    -0.57
     mined
    -0.56
     v
    -0.56
     whit
    -0.56
    POSITIVE LOGITS
    hift
    0.82
    bnb
    0.81
    kward
    0.78
    catentry
    0.77
    ãĥĨãĤ£
    0.76
    APTER
    0.76
    ·
    0.74
    everal
    0.74
    minus
    0.71
     Aires
    0.71
    Act Density 0.186%

    No Known Activations