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

    Neuronpedia

    INDEX
    Explanations

    Arabic prefixes

    np_max-act · gemini-2.0-flash

    The neuron activates on Arabic words related to professions, careers, and vocational/economic terms.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    “To
    -0.07
     LOCATION
    -0.06
     yr
    -0.06
     Tobias
    -0.06
    _PERSON
    -0.06
     pun
    -0.06
     Hudson
    -0.06
     doesn
    -0.06
     cock
    -0.06
     tt
    -0.06
    POSITIVE LOGITS
     الب
    0.08
     الإن
    0.07
     الخ
    0.07
     الأم
    0.07
    /ch
    0.07
     gelen
    0.07
     الموقع
    0.07
    ите
    0.07
     الأسر
    0.06
    الأ
    0.06
    Act Density 0.025%

    No Known Activations