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

    Neuronpedia

    INDEX
    Explanations

    coverage, keep, stay

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ı
    1.25
    한
    1.23
    트
    1.22
    他
    1.16
     dismay
    1.12
    IZ
    1.11
    ătoare
    1.09
     reforzar
    1.09
    어
    1.07
     as
    1.05
    POSITIVE LOGITS
    .
    1.51
    ح
    1.47
    ص
    1.38
    د
    1.37
    л
    1.29
    ä
    1.27
    ీ
    1.25
    </h2>
    1.22
     was
    1.20
    درا
    1.16
    Act Density 0.905%

    No Known Activations