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

    Neuronpedia

    INDEX
    Explanations

    dems

    np_max-act · gemini-2.0-flash

    The neuron fires on subword tokens that are part of proper names or titles (e.g. party names like “Lib Dems” or course names like “Avances en Bioquímica Clínica…”), i.e. named‐entity fragments.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    (Collection
    -0.07
     ص
    -0.07
    (Throwable
    -0.07
    -orders
    -0.06
    .unlink
    -0.06
    (dialog
    -0.06
    фі
    -0.06
    gos
    -0.06
    mir
    -0.06
    .Base
    -0.06
    POSITIVE LOGITS
     buffered
    0.07
     khám
    0.06
    azel
    0.06
     degree
    0.06
     Prosecutor
    0.06
     predicts
    0.06
     illum
    0.06
     ejected
    0.06
     wür
    0.06
    iltere
    0.06
    Act Density 0.469%

    No Known Activations