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

    Neuronpedia

    INDEX
    Explanations

    saw or seen

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    why
    0.40
     зачем
    0.40
    ✿
    0.38
    ﺴ
    0.37
    为何
    0.37
    برای
    0.37
     알아
    0.37
     جهت
    0.36
    PR
    0.36
    気持ち
    0.35
    POSITIVE LOGITS
     saw
    0.98
     seen
    0.95
     gesehen
    0.88
    saw
    0.86
     видел
    0.83
     Seen
    0.82
     Saw
    0.80
    Seen
    0.80
    Saw
    0.78
     sights
    0.77
    Act Density 0.011%

    No Known Activations