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

    Neuronpedia

    INDEX
    Explanations

    phrases related to exceptions or special cases

    oai_token-act-pair · gpt-3.5-turbo

    expressions indicating limitations or constraints

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ammy
    -0.75
    abouts
    -0.69
    ascus
    -0.69
     Started
    -0.67
    eport
    -0.67
    intend
    -0.66
    stanbul
    -0.66
    nas
    -0.65
    nen
    -0.64
    orthern
    -0.64
    POSITIVE LOGITS
     Conversely
    1.23
    Therefore
    1.20
     whereas
    1.14
    Thus
    1.14
     Suppose
    1.12
     Thus
    1.10
     Similarly
    1.10
     Therefore
    1.09
     Consequently
    1.09
     therefore
    1.08
    Act Density 0.576%

    No Known Activations