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

    Neuronpedia

    INDEX
    Explanations

    superlatives or extremes such as "hottest," "safest," "richest," "hardest," and "weakest."

    oai_token-act-pair · gpt-3.5-turbo

    superlative adjectives describing extremes or comparisons

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    skirts
    -0.90
    mun
    -0.89
    bara
    -0.81
    mbuds
    -0.79
    ending
    -0.75
    lim
    -0.75
    bind
    -0.75
    dinand
    -0.74
    cultural
    -0.74
    kas
    -0.74
    POSITIVE LOGITS
     possible
    1.06
     imaginable
    1.04
     conceivable
    1.01
     ever
    0.88
    hest
    0.82
     practicable
    0.77
     thing
    0.74
     strain
    0.72
     contestant
    0.72
     EVER
    0.71
    Act Density 0.047%

    No Known Activations