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

    Neuronpedia

    INDEX
    Explanations

    phrases or sentences emphasizing contrast or contradiction

    oai_token-act-pair · gpt-3.5-turbo

    phrases that express skepticism or doubt

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    hang
    -0.84
    ais
    -0.72
    roxy
    -0.69
    yne
    -0.68
    eki
    -0.67
    iem
    -0.67
    ItemThumbnailImage
    -0.66
    tein
    -0.65
    spin
    -0.65
    MAT
    -0.64
    POSITIVE LOGITS
     ever
    1.02
     bother
    0.83
     EVER
    0.82
     percept
    0.80
     conceivable
    0.80
     imaginable
    0.79
     noticeable
    0.74
     distinguish
    0.74
     believable
    0.72
     bothering
    0.72
    Act Density 0.047%

    No Known Activations