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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating high likelihood or certainty

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    spection
    -0.75
    pointers
    -0.73
    calling
    -0.70
    vati
    -0.64
     Checking
    -0.63
    anooga
    -0.63
    ARDIS
    -0.62
    assed
    -0.62
     Flex
    -0.61
    lication
    -0.61
    POSITIVE LOGITS
     intensify
    1.31
     provoke
    1.19
     worsen
    1.15
     exacerbate
    1.12
     attract
    1.12
     be
    1.11
     become
    1.11
     explode
    1.11
     elicit
    1.08
     offend
    1.05
    Act Density 0.124%

    No Known Activations