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

    Neuronpedia

    INDEX
    Explanations

    Code

    np_max-act · gemini-2.0-flash

    the neuron responds to C++ code identifiers and header names composed of mixed‐case or all‐uppercase tokens (e.g. CamelCase class names and .h filenames).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     toy
    -0.07
     WHY
    -0.07
    .wav
    -0.07
     тем
    -0.07
     مرحله
    -0.06
     WARN
    -0.06
    .domain
    -0.06
     лечения
    -0.06
     kah
    -0.06
     Colour
    -0.06
    POSITIVE LOGITS
    ‚Ì
    0.06
     underestimate
    0.06
    rious
    0.06
    _CHILD
    0.06
     Altern
    0.06
    .And
    0.06
    نوع
    0.06
     家
    0.06
     Mention
    0.06
    BackgroundColor
    0.06
    Act Density 0.069%

    No Known Activations