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

    Neuronpedia

    INDEX
    Explanations

    queries or requests starting with "Can" and followed by a specific action or question

    oai_token-act-pair · gpt-3.5-turbo

    requests or inquiries that start with the word "Can."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     sob
    -0.65
    uria
    -0.61
     fugitive
    -0.61
     Frey
    -0.61
     throat
    -0.60
     residue
    -0.60
     rogue
    -0.59
     hours
    -0.59
     adjud
    -0.58
     heroin
    -0.57
    POSITIVE LOGITS
    Can
    3.51
    CAN
    2.19
     Can
    2.07
    can
    2.00
    Could
    1.70
    Does
    1.64
    Has
    1.62
    Is
    1.53
    Are
    1.48
     CAN
    1.41
    Act Density 0.015%

    No Known Activations