Indexed on: 20 Jul '13Published on: 20 Jul '13Published in: Mathematics - Number Theory
We show a smoothed version of Goldston-Pintz-Yildirim's sifting argument to detect small gaps between primes, which has a higly flexible error term. Our argument is applicable to high dimensional Selberg sieve situations as well, although the relevant details are not stated explicitly. In the present v.2, we have added Appendix in which we made a minor correction to our reasoning following formula (5.14). We are indebted to Professor Terrence Tao for pointing out the necessity of this amendment. The text of the original version has not been changed, except for a correction of the title in the item  of the references.