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

    Neuronpedia

    INDEX
    Explanations

    digits and numbers

    np_max-act · gemini-2.0-flash

    The neuron signals on number tokens and adjacent words specifying numeric quantities (e.g. digit counts or byte lengths).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    Bulletin
    -0.07
     bác
    -0.07
     Sweet
    -0.06
    ，但
    -0.06
     fais
    -0.06
     سرمایه
    -0.06
    .flip
    -0.06
    Suite
    -0.06
     forfe
    -0.06
    (Throwable
    -0.06
    POSITIVE LOGITS
    _Store
    0.06
    mj
    0.06
     Pine
    0.06
     portals
    0.06
    atasets
    0.06
     رض
    0.06
    rant
    0.06
     путем
    0.06
     phủ
    0.06
     власності
    0.06
    Act Density 0.048%

    No Known Activations