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

    Neuronpedia

    INDEX
    Explanations

    Groups of people

    np_max-act · gemini-2.0-flash

    The neuron responds to Cyrillic word fragments—especially common Russian inflectional endings (e.g. “-ую”)—indicating it detects Russian morphological suffixes.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Backup
    -0.07
    Date
    -0.06
    Fd
    -0.06
     Greene
    -0.06
    风险
    -0.06
     hemisphere
    -0.06
    >Status
    -0.06
     CFG
    -0.06
    洛
    -0.06
    ея
    -0.06
    POSITIVE LOGITS
     haft
    0.07
     み
    0.06
     deney
    0.06
     buttons
    0.06
     customers
    0.06
    �始化
    0.06
     chúng
    0.06
    wich
    0.06
     stagn
    0.06
    ้ำหน
    0.06
    Act Density 0.166%

    No Known Activations