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

    Neuronpedia

    INDEX
    Explanations

    generate text or code

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ian
    0.98
     և
    0.96
    arı
    0.95
    étrique
    0.91
    ۔
    0.88
     can
    0.87
    的对象
    0.86
    ղ
    0.84
    с
    0.84
     antérieur
    0.82
    POSITIVE LOGITS
    st
    1.50
    트
    1.38
    ل
    1.36
    सँग
    1.20
    f
    1.12
    is
    1.08
    Z
    1.01
    ר
    0.99
    د
    0.99
    ج
    0.99
    Act Density 0.107%

    No Known Activations