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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating a call to action or directive

    oai_token-act-pair · gpt-3.5-turbo

    references to the phrase "Check out."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    amorph
    -0.74
    uma
    -0.65
    pheus
    -0.64
     bitterness
    -0.61
    ira
    -0.61
     volumes
    -0.61
    pher
    -0.60
     spared
    -0.60
     impart
    -0.60
    channelAvailability
    -0.59
    POSITIVE LOGITS
     Check
    3.40
    Check
    2.64
     check
    2.25
    check
    2.18
     Checks
    2.02
    CHECK
    1.89
     Checking
    1.74
     checks
    1.68
     checking
    1.57
    checking
    1.47
    Act Density 0.012%

    No Known Activations