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

    Neuronpedia

    INDEX
    Explanations

    thousands of

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     و
    1.21
    ی
    1.14
    یت
    1.09
    تی
    0.99
     व
    0.98
     and
    0.96
    و
    0.94
     था
    0.93
     was
    0.93
    was
    0.92
    POSITIVE LOGITS
     thousands
    1.06
    '
    1.04
    (
    1.00
    us
    0.98
     Thousands
    0.96
    الية
    0.88
    \
    0.87
    يرين
    0.86
    -
    0.86
    Z
    0.86
    Act Density 0.017%

    No Known Activations