← All problems

Wasserstein-rate conjecture for Kloosterman sums

Let pp range over primes, let μp\mu_p be the distribution of the normalized Kloosterman sums Kl(a;p)=p−1/2∑x∈Fp×e2πi(ax+x−1)/p\mathrm{Kl}(a;p)=p^{-1/2}\sum_{x\in\mathbb{F}_p^\times}e^{2\pi i(ax+x^{-1})/p} for a∈Fp×a\in\mathbb{F}_p^\times, and let μST\mu_{\mathrm{ST}} be the Sato–Tate law on [−2,2][-2,2] with density 12π4−x2\frac{1}{2\pi}\sqrt{4-x^2}. Prove the conjectured optimal prime-modulus rate Ws(μp,μST)≪sp−1/2W_s(\mu_p,\mu_{\mathrm{ST}})\ll_s p^{-1/2} for every 1≤s≤∞1\le s\le\infty, with an implied constant independent of pp.

References

Progress summary

Refreshed
Open · progress reported

A new preprint improves the known convergence rates for Kloosterman sums, but the conjectured best rate remains unproved.

The problem asks for quantitative convergence of normalized Kloosterman sums to the Sato–Tate law, including the optimal rate suggested by Kowalski and Untrau. The latest work answers their question for every Wasserstein order, but does not establish the conjectured optimal prime-modulus rate.

October 12, 2026 Wasserstein-rate improvement

On October 12, 2026, Amir M. Rahimi's preprint Wasserstein rates of all orders for the Sato-Tate law of Kloosterman sums claims, for 1≤s≤21\le s\le2, Ws≪q−1/3W_s\ll q^{-1/3}; for 2≤s<∞2\le s<\infty, Ws≪q−1/6−1/(3s)W_s\ll q^{-1/6-1/(3s)}; and W∞≪q−1/6W_\infty\ll q^{-1/6}. It improves earlier bounds for s>2s>2 and s=∞s=\infty and gives the s=2s=2 rate via an explicit smoothing argument, but explicitly leaves the conjectured optimal prime-modulus rate unresolved.

Current status (as of October 2026): The all-order Wasserstein bounds are claimed in a new preprint, while the conjectured optimal prime-modulus rate remains open and the advance is unverified.

Sources

Solutions 0

No solutions have been posted yet.