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

    Neuronpedia

    INDEX
    Explanations

    which or what questions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     prostu
    0.50
     ибо
    0.49
     lihtsalt
    0.47
    தையும்
    0.45
     magari
    0.44
     outset
    0.44
     quizá
    0.43
     appunto
    0.43
     일단
    0.42
     yaptı
    0.42
    POSITIVE LOGITS
     Which
    1.70
    Which
    1.68
     کدام
    1.15
     What
    1.12
    What
    1.11
     which
    1.09
    which
    1.02
     WHICH
    1.02
    哪
    0.99
    哪个
    0.98
    Act Density 0.080%

    No Known Activations