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

    Neuronpedia

    INDEX
    Explanations

    adjectives describing people

    oai_token-act-pair · gpt-3.5-turbo

    references to people and their characteristics

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    seys
    -0.68
    ansk
    -0.67
     VIDEOS
    -0.67
    iners
    -0.65
    uld
    -0.64
    uden
    -0.64
    undreds
    -0.64
    çīĪ
    -0.61
    ousands
    -0.61
     Mehran
    -0.61
    POSITIVE LOGITS
     whose
    0.98
     who
    0.94
     whom
    0.89
     nonetheless
    0.86
    nered
    0.84
     indeed
    0.80
     devoid
    0.76
     capable
    0.75
     endowed
    0.74
     possessing
    0.73
    Act Density 0.199%

    No Known Activations