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

    Neuronpedia

    INDEX
    Explanations

    words related to the state of being alive or dead

    oai_token-act-pair · gpt-3.5-turbo

    keywords related to the state or condition of objects or systems

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     reconc
    -0.63
     humble
    -0.61
     frustrations
    -0.56
     wom
    -0.56
     backs
    -0.54
     spoon
    -0.53
     rhy
    -0.53
     Gre
    -0.52
     gull
    -0.52
     skirts
    -0.52
    POSITIVE LOGITS
     enough
    1.27
    izable
    1.20
    isable
    1.13
    enough
    1.11
    ened
    0.97
    locked
    0.97
    rated
    0.95
    iated
    0.92
    able
    0.91
    ised
    0.91
    Act Density 0.282%

    No Known Activations