The generalized arithmetic-progression divisor conjecture

Let nn be a positive integer. The nn-divisor property means that every integer from 11 through nn divides at least one member of a specified product. The generalized arithmetic-progression divisor conjecture. There exist c=O~(logn)c=\widetilde{O}(\log n) integers A1,,AcA_1,\ldots,A_c, each of size O~(nn1/3)\widetilde{O}(n^{n^{1/3}}), and integers m1,,mc0m_1,\ldots,m_c\ge0 such that

1,,ni1,,ic=1m1,,mc(B+i1A1++icAc),1,\ldots,n\mid\prod_{i_1,\ldots,i_c=1}^{m_1,\ldots,m_c}\left(B+i_1A_1+\cdots+i_cA_c\right),

with i=1cmi=O~(n2/3)\prod_{i=1}^{c}m_i=\widetilde{O}(n^{2/3}); equivalently, the index tuples lie in a hypercube of volume O~(n2/3)\widetilde{O}(n^{2/3}).

This generalizes the arithmetic-progression formulation and is proposed as a possible route to improved factorization algorithms. The source says these conjectures are unproved and gives no resolution evidence.

Sources & referencesView supporting material

Primary source

Chris Umans and Siki Wang, “A number-theoretic conjecture implying faster algorithms for polynomial factorization and integer factorization”, arXiv:2511.10851 (2025).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.