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

    Neuronpedia

    INDEX
    Explanations

    number/digit followed by dot or zero

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    на
    0.49
    و
    0.45
    c
    0.44
    and
    0.43
    RO
    0.42
    os
    0.42
    or
    0.41
    as
    0.39
    8
    0.39
    نا
    0.38
    POSITIVE LOGITS
     Ли
    0.34
     Ба
    0.33
     Ви
    0.33
     Анти
    0.33
     {
    0.32
    </td>
    0.32
     Перейти
    0.31
     До
    0.31
     Пре
    0.31
     С
    0.30
    Act Density 0.173%

    No Known Activations