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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating a unique or important role in a narrative

    oai_token-act-pair · gpt-3.5-turbo

    references to "the guy" or "the person" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    mares
    -0.78
    ries
    -0.73
    Animation
    -0.66
    Versions
    -0.66
    cies
    -0.66
    deals
    -0.65
     assumes
    -0.65
     Includes
    -0.64
    amsung
    -0.64
    fn
    -0.63
    POSITIVE LOGITS
     happiest
    1.26
     strongest
    1.10
     smartest
    1.09
     safest
    1.08
     fastest
    1.05
     same
    1.01
     easiest
    1.01
     hardest
    1.00
     richest
    1.00
     closest
    0.95
    Act Density 0.148%

    No Known Activations