Accoring to the Wikipedia article Schrödinger's cat, the answer is yes. I really don't think so. I don't think Schrödinger's cat is a good example because in fact, it's not the case that it will with 100% probability die if and only if an atom decays. Let's instead consider the example of the homogeneous detonation of nitroglycerin where just a few atoms cause an explosive chain reaction. I think there is a way for it to function like it's in a perfect black box. I once read on the internet that for some black holes, it's possible to enter and then avoid reaching the singularity. To ensure that nothing other than the nitroglycerin container itself is affecting the nitroglycerin in any way, let's assume you're watching it from inside such a black hole and the nitroglycerin container is the only thing in the universe outside the black hole. There's no way it's not both exploded and unexploded just because you made a conscious observation because the information about your observation can't escape the black hole to tell its wave function to collapse. I think there's a total misunderstanding of what observation means.
I think this is a simplified model of what would really happen in that situation according to the Copenhagen interpretation. After enough time, a few atoms will happen to get into the right state after their wave functions all collapses and then trigger an explosive chain reaction at only one clear definite time and superposition really only exists at the atomic level. Isn't it only the many worlds interpretation and not the Copenhagen that predicts that the system is in a superposition of exploding at different times? Also, doesn't the many worlds interpretation predict that it is in superposition but not because you are observing it and your own observation puts you into a superposition of observing it explode at different times?