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

    Neuronpedia

    INDEX
    Explanations

    Programming/Log data

    np_max-act · gemini-2.0-flash

    The neuron responds to tokens carrying numerical or technical‐identifier content—dates, version numbers, statistics, measurements, domain names, CVE IDs, durations, and similar numeric/parameter strings.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Lights
    -0.07
     Horde
    -0.06
     Guards
    -0.06
    ávat
    -0.06
     semana
    -0.06
    zilla
    -0.06
    (levels
    -0.06
    控制
    -0.06
    ([('
    -0.06
     damaged
    -0.06
    POSITIVE LOGITS
    maximum
    0.07
     transplant
    0.07
    Nr
    0.07
    :l
    0.07
    IMITIVE
    0.06
     Sporting
    0.06
    :image
    0.06
    _CSR
    0.06
     yacc
    0.06
    еріг
    0.06
    Act Density 0.620%

    No Known Activations