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

    Neuronpedia

    INDEX
    Explanations

    postcard attributes described in detail, with a specific focus on when elements have been written in pencil

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Negative Logits
    <bos>
    -0.90
    sim
    -0.66
     ve
    -0.64
    Já
    -0.64
    Mam
    -0.62
    ellido
    -0.60
     Mam
    -0.60
     Hub
    -0.60
    Hub
    -0.60
    mainAxisSize
    -0.59
    POSITIVE LOGITS
     pencil
    2.68
     Pencil
    2.42
     pencils
    2.34
    Pencil
    2.29
    pencil
    1.98
     crayon
    1.66
     impra
    1.55
     crayons
    1.53
     stockholm
    1.49
     eiffel
    1.49
    Act Density 0.146%

    No Known Activations