Erdős Problem #164 — The primitive-set reciprocal logarithmic sum

About 40 years old · traced to

A set A⊆NA\subseteq\mathbb N is primitive if no member of AA divides another. Prove that for every primitive set AA whose elements are all at least 22,

∑a∈A1alog⁡a≤∑p prime1plog⁡p.\sum_{a\in A}\frac{1}{a\log a}\leq \sum_{p\ \mathrm{prime}}\frac{1}{p\log p}.
Equivalent formulations 2Other wordings

Other statements of this same problem, merged from separate entries. Each is equivalent to the statement above — proving any one settles them all.

  1. Erdős primitive-set inequality

    Among primitive sets of integers greater than one, is the weighted reciprocal sum from Erdős's conjecture maximized by the primes?

    source: Open Math Problems Claimed to Be Solved with AI

  2. Erdős conjecture for primitive sets

    A subset of the integers larger than 11 is primitive if no member divides another. Define f(a)=1/(alog⁡a)f(a)=1/(a\log a) and f(A)=∑a∈Af(a)f(A)=\sum_{a\in A}f(a). Let P(A)\mathcal{P}(A) denote the set of primes that divide some member of AA. Erdős conjecture. For any primitive set AA, we have

    f(A)≤f(P(A)).f(A) \le f(\mathcal{P}(A)).

    Erdős proved that f(A)f(A) is universally bounded over primitive sets, and conjectured that this bound is attained by the set of prime numbers. The conjecture remains open, although progress is known in certain cases.

    source: Jared Duker Lichtman, Greg Martin and Carl Pomerance, “Primes in prime number races”, arXiv:1809.03033 (2019).

References

Progress summary

Refreshed
Claimed solved

The conjecture that the primes give the largest weighted reciprocal sum over primitive sets has been proved, although a related technical question about the prime 22 remains open.

The problem asks whether every primitive set AA satisfies f(A)≤f(P)f(A)\leq f(\mathcal P), where f(A)=∑a∈A1/(alog⁡a)f(A)=\sum_{a\in A}1/(a\log a) and P\mathcal P is the primes. Erdős proved uniform boundedness in 19351935 and posed the maximization question in the 19801980s.

Known results

  • Erdős and Zhang, 19931993: f(A)<1.84f(A)<1.84 for every primitive set.
  • Zhang, 19911991: the prime-maximization statement for sets of integers with at most four prime factors counted with multiplicity.
  • Lichtman and Pomerance, 20192019: f(A)<eγ=1.781…f(A)<e^\gamma=1.781\ldots.
  • Lichtman, 20232023: the original conjecture proved.

Proofs and formalization, 2022–2026

A 20222022 arXiv paper proves f(A)≤f(P)=1.6366…f(A)\leq f(\mathcal P)=1.6366\ldots, establishing the conjecture. The result is treated as settled in later work; an alternative proof by Alexeev, Barreto, Li, Lichtman, Price, Shah, Tang, and Tao appeared in 20262026, with a related version formalized in Lean by Math Inc.

Current status (as of May 2026): The Erdős primitive-set inequality is proved; whether 22 is Erdős strong remains open as an auxiliary question.

Sources

Solutions 0

No solutions have been posted yet.