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

    Neuronpedia

    INDEX
    Explanations

    patch test on a small

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Fantasy
    0.47
     도
    0.46
    akk
    0.46
    ະຍ
    0.46
     가
    0.45
    बास
    0.45
    ное
    0.45
     신
    0.43
    лым
    0.42
     kip
    0.42
    POSITIVE LOGITS
     poured
    0.54
    );
    0.53
    ).
    0.52
    utig
    0.49
    .).
    0.48
    ि
    0.48
     konular
    0.48
    Questa
    0.48
    )।
    0.47
     morte
    0.47
    Act Density 0.000%

    No Known Activations