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

    Neuronpedia

    INDEX
    Explanations

    Probability calculations

    np_max-act · gemini-2.0-flash

    The neuron detects the pattern of a numeric count immediately followed by a single letter (e.g. “2 t” or “3 f”), as used in the probability‐of‐picking formulations.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    .GraphicsUnit
    -0.06
    @GetMapping
    -0.06
    never
    -0.06
    eil
    -0.06
    datal
    -0.06
    664
    -0.06
    erry
    -0.06
    .Shapes
    -0.06
    озможно
    -0.05
     acı
    -0.05
    POSITIVE LOGITS
     occur
    0.07
     les
    0.07
     products
    0.07
    \↵
    0.07
     влад
    0.06
     Come
    0.06
    FE
    0.06
     Les
    0.06
     editar
    0.06
     PSI
    0.06
    Act Density 0.003%

    No Known Activations