Admissibility of the constant 1 in the weighted large sieve

Determine whether the constant 1 is admissible in place of the current coefficient in the weighted large sieve inequality, thereby improving the coefficient appearing in the denominator of the associated sieve bound.

Background

The paper derives a weighted large sieve inequality whose denominator contains the term c q(q+Q), with c=0.898678… in the paper’s formulation, and compares this with the classical coefficient 3/2 in the Montgomery–Vaughan result. The discussion notes that E. Preissmann improved the coefficient 3/2 to approximately 1.315 and states that the constant 1 is conjectured to be admissible.

Establishing admissibility of the constant 1 would yield a stronger weighted large sieve inequality and, according to the paper, would improve the numerical constant in the resulting upper bound for the number of twin primes in an interval.

References

E.~Preissmann in improved this $3/2$ to~$\sqrt{1+\frac23\sqrt{6/5}=1.315\dots$ and this appears to be the last improvement on this matter, though the constant~$1$ is conjectured to be admissible.

— The weighted large sieve through Parseval  (2609.25885 - Ramaré, 22 Sep 2026) in Section 1, subsection “A first general result”