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

    Neuronpedia

    INDEX
    Explanations

    Code symbols

    np_max-act · gemini-2.0-flash

    This neuron detects the opening of code blocks in source code, especially the “{” that begins class, function, or other block definitions.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     zi
    -0.07
    nen
    -0.07
     QUAL
    -0.06
    (variable
    -0.06
    ício
    -0.06
    лаг
    -0.06
    >NN
    -0.06
     buckets
    -0.06
     torch
    -0.06
    -small
    -0.06
    POSITIVE LOGITS
    supports
    0.07
    .DESC
    0.07
     参数
    0.06
     "");
    0.06
     مسجد
    0.06
    angep
    0.06
     ifdef
    0.06
     hodně
    0.06
    _ylim
    0.06
    除了
    0.06
    Act Density 0.035%

    No Known Activations