---

Claim. The question of Problem 475 has the answer yes for every prime $p\ge p_0$, for some threshold $p_0$ that no source makes explicit: every $A\subseteq\mathbb F_p\setminus\{0\}$ has an ordering whose partial sums are distinct. What remains is a finite check, the primes below $p_0$. The claim is stated in H. T. Pham and L. Sauermann, On Graham's rearrangement conjecture (arXiv:2602.15797, 17 February 2026), whose Theorem 1.2 gives, for any fixed $0<\alpha<1$, a constant $C_\alpha$ such that every $S\subseteq\mathbb Z_p\setminus\{0\}$ with $C_\alpha\le|S|\le p^{1-\alpha}$ has a valid ordering, by an anticoncentration bound for the sum of a random subset and local repair of a random ordering at each zero-sum segment, and which says (p. 2) that together with the earlier results this settles the conjecture for all sufficiently large primes. The chain it completes has four ranges. Small $t$: Theorem 1.2 of Bedert and Kravitz (Israel J. Math. 273 (2026); claim page), $t\le e^{c(\log p)^{1/4}}$ for every constant $c>0$ and every large prime, by a structure theorem into dissociated sets and a rectifiable remainder; it extends Kravitz's $t\le\log p/\log\log p$ for every prime, and Costa and Della Fiore's Theorem 1.3 (a 2026 preprint) later widened the range to $t\le e^{c(\log p)^{1/3}}$ for some $c>0$. Medium $t$: Pham and Sauermann's theorem. Large $t$: Theorem 1.4 of Bedert, Bucić, Kravitz, Montgomery and Müyesser (arXiv:2508.18254), an absolute $c>0$ such that $t\ge p^{1-c}$ suffices in every finite group, by absorption and a regularity decomposition of Cayley graphs. Very large $t$: their Theorem 7.1, $t\ge p-p^{1-\gamma}$ for large $p$, derived in its Appendix A from the random Hall--Paige machinery of Müyesser and Pokrovskiy (Invent. Math. 240 (2025)): their Lemma 6.22 and the method of their Theorem 6.9. Fixing $\alpha\le c$, the ranges overlap once $p$ is large enough that $e^{c'(\log p)^{1/4}}\ge C_\alpha$, so every size $t$ is covered for $p\ge p_0$.

Covers. Every prime $p\ge p_0$, with $p_0$ unstated: the four papers say only "large prime", "$C_\alpha$", "absolute constant $c$" and "sufficiently large $N$", so the finite check has no known extent. For every prime the statement is known for $t\le12$, by Proposition 4.2 of Costa and Pellegrini (Arch. Math. 115 (2020)), for every set of size $p-2$ or $p-1$, by Bode and Harborth (Discrete Math. 299 (2005)), and for every $(p-3)$-subset with nonzero sum, by Hicks, Ollis and Schmitt (J. Combin. Des. 27 (2019)); these every-prime results have their own pages (Costa and Pellegrini, Bode and Harborth, Hicks, Ollis and Schmitt and Kravitz). The check that would close the problem is therefore the sizes $13\le t\le p-4$ and the zero-sum $(p-3)$-subsets for the finitely many primes below $p_0$, and nothing bounds that set.

Depends on. For the small range, Bedert and Kravitz's claim page; the other links have no claim pages and are the papers' own, filed on their library result pages.

Standing. Claimed. Not reviewed: the site's curator records in the problem page's commentary that the conjecture is proved for all sufficiently large primes as a consequence of four kinds of result and credits each paper with its range (page last edited 5 March 2026; empty proof-claim tab), and the community database lists the problem as decidable, as of its last update on 23 February 2026; but the site's label DECIDABLE leaves the problem open and settles no part of it, so the commentary records the chain without accepting it as a solution. Not refereed as a whole: the papers of Bedert and Kravitz and of Müyesser and Pokrovskiy are refereed, but Pham and Sauermann's and Bedert, Bucić, Kravitz, Montgomery and Müyesser's are preprints with no journal record on 2026-09-18, so two of the four links carry the preprint qualification. Not formalized: no Lean statement or proof of the problem existed in the catalog on 2026-09-18. Read depth: claims checked on the library result pages (2026-09-18); no proof in the chain checked.