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

    Neuronpedia

    INDEX
    Explanations

    scientific findings and methods

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     所以
    0.29
     nějak
    0.28
     nên
    0.28
     sbParams
    0.27
    所以我
    0.26
     그대로
    0.26
    طيني
    0.26
    もう少し
    0.26
     Bakın
    0.26
    があるので
    0.25
    POSITIVE LOGITS
    We
    0.36
    Experimental
    0.31
    Recogn
    0.30
    A
    0.30
    Using
    0.30
     We
    0.29
    Pre
    0.29
    First
    0.29
    In
    0.29
    The
    0.28
    Act Density 0.003%

    No Known Activations