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

    Neuronpedia

    INDEX
    Explanations

    boxes and rectangles

    np_max-act · gemini-2.0-flash

    The neuron activates on code drawing rectangle shapes—e.g. calls to “pygame.draw.rect” or references to drawing a rectangle.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     prized
    -0.07
     オ
    -0.07
     multiplication
    -0.06
     termination
    -0.06
    IDX
    -0.06
    ứt
    -0.06
     squadron
    -0.06
     Starbucks
    -0.06
    (row
    -0.06
    jší
    -0.06
    POSITIVE LOGITS
    /live
    0.07
     '>'
    0.06
     adipiscing
    0.06
     Jas
    0.06
     \↵
    0.06
    Rectangle
    0.06
     bacheca
    0.06
    .getServer
    0.06
    sn
    0.06
     memberId
    0.06
    Act Density 0.038%

    No Known Activations