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

    Neuronpedia

    INDEX
    Explanations

    ,

    np_max-act · gemini-2.0-flash

    This neuron responds to capitalized name‐like tokens—i.e. proper nouns (people, organizations, dates, etc.).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     scand
    -0.08
    iyan
    -0.07
     розрах
    -0.06
    orage
    -0.06
    णन
    -0.06
     onChanged
    -0.06
    aa
    -0.06
    ると
    -0.06
     UNU
    -0.06
    962
    -0.06
    POSITIVE LOGITS
     Wrestling
    0.06
    .Video
    0.06
     Beit
    0.06
    ;
    0.06
     MP
    0.06
    modelo
    0.06
     looking
    0.06
    "While
    0.06
    .Device
    0.06
    _regs
    0.06
    Act Density 0.035%

    No Known Activations