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

    Neuronpedia

    INDEX
    Explanations

    Sieve of Eratosthenes

    np_max-act · gemini-2.0-flash

    This neuron is sensitive to the special control tokens used by the system (e.g. start/end headers or end-of-text markers) rather than natural language words.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     ошиб
    -0.07
    liche
    -0.07
     Canvas
    -0.06
     prakt
    -0.06
    -spe
    -0.06
     Word
    -0.06
    Histogram
    -0.06
    deme
    -0.06
     Cartoon
    -0.06
    _with
    -0.06
    POSITIVE LOGITS
    ingle
    0.06
     عمومی
    0.06
    ゛
    0.06
    Protected
    0.06
     الإ
    0.06
     shuttle
    0.06
     eventually
    0.06
    ,i
    0.06
     الکتر
    0.06
    §Ã
    0.06
    Act Density 0.014%

    No Known Activations