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

    Neuronpedia

    INDEX
    Explanations

    variables in math and code

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    y
    3.20
    r
    3.16
    at
    3.09
    on
    2.99
    in
    2.98
    u
    2.90
    l
    2.85
    m
    2.83
    ل
    2.82
    h
    2.77
    POSITIVE LOGITS
    Какой
    1.59
    름
    1.42
     dàng
    1.37
     て
    1.35
    clickView
    1.29
     rays
    1.28
    Какие
    1.27
    ߋ
    1.24
     t
    1.24
    Полу
    1.22
    Act Density 0.731%

    No Known Activations