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

    Neuronpedia

    INDEX
    Explanations

    Arabic phrases and country names

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     aprire
    0.34
     cornerback
    0.31
     huit
    0.31
     quirky
    0.30
     cheeky
    0.30
     problemlos
    0.30
     deux
    0.29
     voiture
    0.29
     kaikki
    0.29
     tiež
    0.29
    POSITIVE LOGITS
    ‌
    0.26
    ית
    0.25
    ِ
    0.25
    نا
    0.25
    5
    0.25
    ี
    0.24
     ا
    0.24
    ُ
    0.24
     הר
    0.23
    ـــ
    0.23
    Act Density 0.017%

    No Known Activations