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

    Neuronpedia

    INDEX
    Explanations

    references to circular objects or concepts

    oai_token-act-pair · gpt-3.5-turbo

    references to shapes, specifically circles

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    bid
    -0.82
    orthy
    -0.73
    offer
    -0.67
     Reloaded
    -0.67
    pp
    -0.64
    graded
    -0.64
    cedented
    -0.62
    fet
    -0.62
    added
    -0.62
    FD
    -0.62
    POSITIVE LOGITS
     circle
    4.00
    circle
    2.90
     circles
    2.63
     Circle
    2.40
     sphere
    1.58
     triangle
    1.45
     rectangle
    1.42
     circled
    1.42
     circling
    1.37
     radius
    1.33
    Act Density 0.023%

    No Known Activations