Erdős Problem 260

∀a:N→Z, ∀s∈R, (StrictMono⁡(a) ∧ Tendsto⁡ ⁣(n↦a(n)n, atTop⁡, atTop⁡) ∧ HasSum⁡ ⁣(n↦2−a(n)1, s)) ⟹ Irrational⁡(s).\forall a:\mathbb{N}\to\mathbb{Z},\ \forall s\in\mathbb{R},\ \Bigl(\operatorname{StrictMono}(a)\ \land\ \operatorname{Tendsto}\!\left(n\mapsto\frac{a(n)}{n},\ \operatorname{atTop},\ \operatorname{atTop}\right)\ \land\ \operatorname{HasSum}\!\left(n\mapsto\frac{2^{-a(n)}}{\phantom{1}},\ s\right)\Bigr)\ \Longrightarrow\ \operatorname{Irrational}(s).

References

Primary source

GitHub

Additional references

Progress summary

Refreshed
Claimed solved

A June 2026 paper claims to settle the problem, and a related Lean project now claims a complete formalization, but neither claim has independent validation.

Erdős Problem 260 asks whether every sufficiently sparse binary expansion of the form ∑n≥12−an\sum_{n\ge1}2^{-a_n}, with an/n→∞a_n/n\to\infty, is irrational. A June 2026 preprint claims an affirmative proof.

June–August 2026 claimed proof and formalization

The June preprint claims a stronger positive-density theorem for the support of a rational binary expansion and derives the stated irrationality result. It reports that the associated Lean project covered only part of the stopping-time argument and finite checks; by August, the repository itself claimed a complete Lean proof without placeholders or project-level axioms. The mathematical proof and formalization remain unverified independently.

Current status (as of August 2026): A preprint claims the irrationality theorem and the repository claims a complete Lean formalization, but independent validation of both the argument and its formalization is still absent.

Sources

Solutions 0

No solutions have been posted yet.