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

    Neuronpedia

    INDEX
    Explanations

    search

    np_max-act · gemini-2.0-flash

    The neuron activates on tokens referring to “find”/search operations (e.g. find, findIndex, indexOf or the Russian иск- root), so it detects mentions of searching or indexing elements in a collection.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Lun
    -0.06
    modelo
    -0.06
    urses
    -0.06
    .photos
    -0.06
    LOGGER
    -0.06
    ys
    -0.06
    ksi
    -0.06
    (cuda
    -0.06
    ysa
    -0.06
    Cs
    -0.06
    POSITIVE LOGITS
     certs
    0.07
    .HtmlControls
    0.07
     غرب
    0.07
    plots
    0.06
     navigation
    0.06
    ifik
    0.06
     cheers
    0.06
    QRST
    0.06
    \Lib
    0.06
    ίναι
    0.06
    Act Density 0.016%

    No Known Activations