부산대학교 도서관

In this document we address an error discovered in the ensemble generation for our calculation of the $I=0$ $K\to\pi\pi$ amplitude (Phys. Rev. Lett. 115, 212001 (2015), arXiv:1505.07863) whereby the same random numbers were used for the two independent quark flavors, resulting in small but measurable correlations between gauge observables separated by 12 units in the y-direction. We conclude that the effects of this error are negligible compared to the overall errors on our calculation.
Comment: 2 pages, 1 figure